Method and system for cleaning a temperature measurement optical window, and semiconductor process apparatus

Abstract

The application provides a temperature measurement optical window pollution processing method and system and a semiconductor process equipment, relates to the technical field of semiconductor detection, and respectively acquires temperature measurement main light path data, auxiliary light path data and process operation data, determines a main light path estimation result according to the temperature measurement main light path data, multiplexes an original temperature measurement main light path to realize online preliminary identification of a window deposition state, supplements actual transmission performance of the temperature measurement optical window according to the auxiliary light path data to improve the reliability of deposition degree judgment, performs multi-parameter Monte Carlo inversion when preset inversion conditions are met, obtains more accurate deposition thickness and deposition optical parameters when complex deposition or conventional judgment is uncertain, finally performs multi-source comprehensive processing to determine deposition state information of the temperature measurement optical window, and determines a pollution processing strategy according to the deposition state information and process stage information, reduces the risk of misjudgment of a single data source, guarantees the accuracy of temperature measurement, and reduces the interference of cleaning behavior on epitaxial process.

Description

Technical Field

[0001] This application relates to the field of semiconductor detection technology, and in particular to a method, system, and semiconductor process equipment for treating contamination of a temperature-measuring optical window. Background Technology

[0002] Semiconductor epitaxial processes such as Metal-Organic Chemical Vapor Deposition (MOCVD) typically require high temperature, low pressure, or specific atmospheric conditions. During epitaxial growth, the temperature within the reaction chamber directly affects the thickness, composition, crystal quality, and device performance of the epitaxial layer. To achieve real-time monitoring of the temperature within the reaction chamber, existing equipment usually employs non-contact temperature measurement methods such as infrared thermometry, dual-color thermometry, or colorimetric thermometry. These methods typically require a temperature-measuring optical window on the reaction chamber, through which the radiation signal from the chamber is transmitted to the external main temperature-measuring optical path, where the corresponding temperature signal is obtained by the temperature measuring instrument.

[0003] However, during epitaxial processes such as MOCVD, metal-organic sources, reaction byproducts, particulate matter, and intracavity volatiles can easily deposit on the surface of the temperature-sensing optical window. As the deposited layer gradually thickens, the transmittance, reflectance, and spectral response characteristics of the temperature-sensing optical window change, leading to attenuation or distortion of the radiation intensity signal received by the main temperature-sensing optical path. For two-color or colorimetric temperature measurement, the degree of attenuation may not be consistent at different wavelengths. Window deposition not only reduces the intensity of the temperature measurement signal but may also change the ratio between signals of different wavelengths, causing drift in the temperature measurement results and affecting the control accuracy of the epitaxial process.

[0004] In traditional techniques, the transmittance of a window is typically measured using an auxiliary light source to determine whether the temperature-measuring optical window is contaminated. However, a single transmittance measurement usually only reflects the transmission change at a specific auxiliary wavelength, making it difficult to accurately distinguish between different deposition types such as absorptive deposition, transparent thin film deposition, and complex interference deposition. When a transparent thin film or multilayer complex deposition forms on the window surface, the transmittance change may be affected by multiple factors, including thin film interference, refractive index changes, extinction coefficient changes, and surface roughness. Relying solely on a single measurement data point or empirical threshold can affect detection accuracy and easily lead to misjudgments. Summary of the Invention

[0005] The technical problem solved by this application is to provide a method, system and semiconductor process equipment for treating contamination of a temperature measuring optical window, so as to overcome the defects of traditional technology that uses a single transmittance to judge the contamination status of the window, which is easily affected by light source fluctuations, thermal radiation background and thin film interference, making it difficult to accurately distinguish the deposition type and deposition degree, resulting in insufficient reliability of contamination judgment results.

[0006] In a first aspect, this application proposes a method for treating contamination in a temperature-sensing optical window, applicable to semiconductor process equipment including a temperature-sensing optical window, the method comprising: Acquire temperature measurement main optical path data, auxiliary optical path data, and process operation data; the process operation data includes process stage information. Based on the temperature measurement main optical path data, the deposition index is calculated, and based on the deposition index, the main optical path estimation result is determined; the main optical path estimation result includes the thickness estimation result and the type prediction result; Based on the auxiliary optical path data, the transmittance data is calculated; If at least one of the deposition index and the transmittance data satisfies a preset inversion condition, a multi-parameter Monte Carlo inversion is performed based on the temperature measurement main optical path data and the auxiliary optical path data to obtain the inversion result; the inversion result includes the inverted deposition thickness and the inverted deposition optical parameters. Based on the main optical path estimation results, the transmittance data, and the inversion results, multi-source integrated processing is performed to determine the deposition state information of the thermometric optical window; wherein, the deposition state information includes at least one of fused deposition thickness, fused deposition optical parameters, and deposition type; Based on the deposition state information and the process stage information, a contamination treatment strategy is determined; the contamination treatment strategy includes at least one of a temperature compensation strategy and an in-situ cleaning strategy.

[0007] In one embodiment, the temperature measurement main optical path data includes a first radiation intensity signal corresponding to a first temperature measurement wavelength and a second radiation intensity signal corresponding to a second temperature measurement wavelength; The deposition index is calculated based on the temperature measurement main optical path data, including: The original temperature value is calculated based on the first radiation intensity signal and the second radiation intensity signal; Calculate the original intensity ratio at the current moment based on the first radiation intensity signal and the second radiation intensity signal; Based on the original temperature measurement value, the corresponding cleanliness intensity ratio is determined from the pre-established cleanliness baseline relationship; the cleanliness baseline relationship is used to characterize the correspondence between different temperature measurement values ​​and the corresponding cleanliness intensity ratio under clean temperature measurement optical window conditions; Based on the original strength ratio and the clean strength ratio, determine the relative deviation information; The relative deviation information is filtered to obtain the deposition index.

[0008] In one embodiment, determining the main optical path estimation result based on the deposition index includes: The variation characteristics of the sedimentation index are determined based on the sedimentation index at different times; Based on the deposition index and the variation characteristics of the deposition index, a type prediction result is determined; wherein, the type prediction result includes at least one of absorptive deposition, transparent thin film deposition, and complex deposition; If the predicted type is absorptive deposition, the thickness estimation result is determined based on the deposition index and the pre-established empirical model of deposition thickness. The type prediction result and / or the thickness estimation result are determined as the main optical path estimation result.

[0009] In one embodiment, the auxiliary optical path data includes a transmission signal acquired after the auxiliary monitoring beam passes through the temperature measuring optical window, and a reference signal used to characterize the output intensity of the auxiliary monitoring light source. The step of calculating transmittance data based on the auxiliary optical path data includes: The transmitted signal and the reference signal are preprocessed to obtain the transmitted signal value and the reference signal value; Acquire the reference transmission signal value and reference reference signal value under clean conditions; The transmittance data is calculated based on the transmitted signal value, the reference signal value, the baseline transmitted signal value, and the baseline reference signal value, and its expression is:

[0010] in, For transmittance data, This is the transmitted signal value. As a reference signal value, As the reference transmission signal value, This is the reference signal value.

[0011] In one embodiment, the method for determining whether the preset inversion conditions are met includes: If at least one of the following conditions is met: the magnitude of change of the deposition index is greater than a preset magnitude threshold, the deposition index oscillates a preset number of times within a preset time period, the model uncertainty corresponding to the deposition index is greater than a preset uncertainty threshold, or the deviation between the transmittance data and the transmittance predicted based on the temperature measurement main optical path data is greater than a preset deviation threshold, then the preset inversion conditions are considered to be met.

[0012] In one embodiment, when at least one of the deposition index and the transmittance data satisfies a preset inversion condition, a multi-parameter Monte Carlo inversion is performed based on the temperature measurement main optical path data and the auxiliary optical path data to obtain the inversion result, including: Under the condition of satisfying the preset inversion, the first transmittance data at at least one temperature measurement wavelength is determined according to the temperature measurement main optical path data, and the second transmittance data at the auxiliary wavelength is determined according to the auxiliary optical path data. After combination, the measured transmittance data at multiple wavelengths are obtained. Using deposition thickness and deposition optical parameters as inversion parameters, a multi-parameter inversion model is constructed to characterize the deposition layer on the surface of the thermometric optical window; wherein, the deposition optical parameters include refractive index and extinction coefficient; Monte Carlo photon transmission simulation was performed based on the multi-parameter inversion model to obtain theoretical transmittance data corresponding to different inversion parameters. Based on the degree of matching between the measured transmittance data and the theoretical transmittance data, a likelihood function is constructed, and Bayesian sampling is performed on the parameters to be inverted based on the likelihood function to obtain the inversion result.

[0013] In one embodiment, the step of performing Bayesian sampling on the parameters to be inverted based on the likelihood function to obtain the inversion result includes: A prior distribution is set for the parameters to be inverted; a candidate parameter group is generated based on the prior distribution; the acceptance probability of the candidate parameter group is determined based on the difference between the theoretical transmittance data and the measured transmittance data corresponding to the candidate parameter group; the candidate parameter group is iteratively sampled according to the acceptance probability; and the inversion result is determined based on the sampling results after the convergence condition is met.

[0014] In one embodiment, the step of performing multi-source synthesis processing based on the main optical path estimation result, the transmittance data, and the inversion result to determine the deposition state information of the thermometric optical window includes: Based on the transmittance data, the auxiliary optical path estimation result is determined; the auxiliary optical path estimation result includes the auxiliary deposition thickness and the auxiliary deposition optical parameters; Based on the main optical path estimation result, the auxiliary optical path estimation result, and the inversion result, a thickness fusion process is performed to obtain the fused deposition thickness; Based on the transmittance data and the inversion results, the optical parameters for fusion deposition are determined; The type prediction result is corrected based on the variation characteristics of the fused deposition optical parameters and the deposition index to obtain the deposition type; the variation characteristics are determined based on the deposition index at different times.

[0015] In one embodiment, the step of performing thickness fusion processing based on the main optical path estimation result, the auxiliary optical path estimation result, and the inversion result to obtain the fused deposition thickness includes: Based on the main optical path estimation result, the auxiliary optical path estimation result, and the reliability of the inversion result, their respective fusion weights are determined; wherein, if the inversion result does not exist or the reliability of the inversion result does not meet the preset reliability condition, the fusion weight corresponding to the inversion result is set to zero; The thickness fusion process is performed using the fusion weight to obtain the fused deposition thickness.

[0016] In one embodiment, determining the fusion deposition optical parameters based on the transmittance data and the inversion result includes: Based on the transmittance data and the preset thin film optical model, parameter inversion is performed to determine auxiliary optical parameters that match the theoretical transmittance with the transmittance data; If the inversion result is not available or the inversion result does not meet the preset confidence level, the auxiliary optical parameter will be determined as the fusion deposition optical parameter. If the inversion results meet the preset confidence level, the inversion optical parameters in the inversion results are determined as the fusion deposition optical parameters.

[0017] In one embodiment, determining the contamination treatment strategy based on the deposition state information and the process stage information includes: The deposition degree of the thermometric optical window is determined based on the fusion deposition thickness and the deposition type. If the deposition level reaches the temperature compensation threshold but does not reach the cleaning threshold, it is determined that the contamination treatment strategy includes a temperature compensation strategy. When the deposition level reaches the cleaning threshold and the process stage information indicates that the process is currently in a cleanable stage, the contamination treatment strategy is determined to include an in-situ cleaning strategy. If the deposition level reaches the cleaning threshold and the process stage information indicates that the process is not currently in a cleanable stage, the contamination treatment strategy is determined to include a temperature compensation strategy.

[0018] In one embodiment, the temperature compensation strategy includes: The temperature compensation amount is determined based on the mapping model between the deposition state information and the temperature deviation; The original temperature value output by the main optical path of temperature measurement is corrected according to the temperature compensation amount to obtain the corrected temperature value. The mapping model is a nonlinear model established based on calibration data. The nonlinear model is used to characterize the temperature measurement deviation corresponding to different deposition state information. The nonlinear model includes at least one of a polynomial model, a lookup table model, or a piecewise fitting model.

[0019] In one embodiment, the in-situ cleaning strategy includes: The temperature-measuring optical window is scanned at multiple points to generate a contamination distribution map; The laser scanning path and the laser dwell time and / or scanning speed corresponding to each scanning area are determined according to the pollution distribution map. The area with a pollution level greater than a preset threshold is denoted as the first area, and the area with a pollution level less than or equal to the preset threshold is denoted as the second area. The laser dwell time corresponding to the first area is greater than the laser dwell time corresponding to the second area, and / or the scanning speed corresponding to the first area is less than the scanning speed corresponding to the second area. During the cleaning process, the auxiliary optical path data is acquired in real time, and the real-time transmittance data of the current area is determined based on the auxiliary optical path data. When the real-time transmittance data recovers to the target transmittance, the subsequent cleaning of the current area is stopped or skipped.

[0020] In one embodiment, the process operating data further includes reaction chamber pressure and pressure change rate; the method further includes: During the operation of the semiconductor process equipment, a protective gas curtain is provided to the temperature-measuring optical window according to a target purge flow rate; The target purging flow rate is obtained by dynamically adjusting the baseline purging flow rate based on the reaction chamber pressure, the pressure change rate, and the process stage information, so as to match the protective gas curtain with the current process state.

[0021] In one embodiment, the method further includes: After the temperature measuring optical window is cleaned in situ, the updated transmittance benchmark value of the temperature measuring optical window in the clean state is recorded. Based on the changing trends of multiple updated transmittance benchmark values, it is determined whether the temperature measuring optical window has experienced substrate aging or irreversible damage. If it is determined that the temperature-measuring optical window has experienced substrate aging or irreversible damage, the temperature compensation threshold, cleaning threshold, and / or the reference parameters used to calculate the transmittance data are adjusted, and a maintenance warning is output.

[0022] Secondly, this application proposes a contamination treatment system for a temperature-sensing optical window, disposed within a semiconductor process apparatus including the temperature-sensing optical window, the system comprising: The acquisition module is used to acquire temperature measurement main optical path data, auxiliary optical path data, and process operation data; the process operation data includes process stage information. The calculation module is used to calculate the deposition index based on the temperature measurement main optical path data, and determine the main optical path estimation result based on the deposition index; the main optical path estimation result includes the thickness estimation result and the type prediction result; and to calculate the transmittance data based on the auxiliary optical path data. The inversion module is used to perform multi-parameter Monte Carlo inversion based on the temperature measurement main optical path data and the auxiliary optical path data, under the condition that at least one of the deposition index and the transmittance data meets a preset inversion condition, to obtain the inversion result; the inversion result includes the inverted deposition thickness and the inverted deposition optical parameters; The integrated processing module is used to perform multi-source integrated processing based on the main optical path estimation result, the transmittance data, and the inversion result to determine the deposition state information of the thermometric optical window; wherein, the deposition state information includes at least one of fused deposition thickness, fused deposition optical parameters, and deposition type; The strategy determination module is used to determine a contamination treatment strategy based on the deposition state information and the process stage information; the contamination treatment strategy includes at least one of a temperature compensation strategy and an in-situ cleaning strategy.

[0023] Thirdly, this application proposes a semiconductor process apparatus, the apparatus comprising: A reaction chamber is used to provide an environment for semiconductor epitaxial processes; An optical window assembly includes a temperature-measuring optical window, which is disposed at the temperature-measuring channel of the reaction chamber and is used to provide a light signal transmission channel. The temperature measurement main optical path assembly is used to receive the radiation signal in the reaction cavity transmitted through the temperature measurement optical window and generate temperature measurement main optical path data. An auxiliary optical path assembly is used to emit an auxiliary monitoring beam and collect auxiliary optical path data obtained after the auxiliary monitoring beam passes through the temperature measuring optical window; A purge assembly, disposed on one side of the temperature measuring optical window, is used to provide a protective air curtain to the temperature measuring optical window; A cleaning module is used to perform in-situ cleaning of the temperature-measuring optical window based on a contamination treatment strategy. The contamination treatment system for the temperature measurement optical window as described in the second aspect is used to determine the contamination treatment strategy based on the temperature measurement main optical path data, auxiliary optical path data, and process operation data.

[0024] Fourthly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method of the first aspect.

[0025] Fifthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the method of the first aspect.

[0026] Sixthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the method in the first aspect.

[0027] The above-mentioned method, system, and semiconductor process equipment for treating contamination of the temperature-measuring optical window have at least the following advantages: This application acquires temperature measurement main optical path data, auxiliary optical path data, and process operation data. Based on the temperature measurement main optical path data, it calculates the deposition index and determines the main optical path estimation result, reusing the original temperature measurement main optical path to achieve online preliminary identification of the window deposition state. Then, based on the auxiliary optical path data, it calculates transmittance data to supplement and characterize the actual transmittance performance of the temperature measurement optical window, improving the reliability of the deposition degree judgment. When preset inversion conditions are met, multi-parameter Monte Carlo inversion is performed to obtain more accurate deposition thickness and deposition optical parameters in complex deposition or when conventional judgments are uncertain. Finally, by comprehensively processing the main optical path estimation result, transmittance data, and inversion result from multiple sources, the deposition state information of the temperature measurement optical window is determined. Based on the deposition state information and the process stage information of the process operation data, a contamination treatment strategy is determined, which can reduce the risk of misjudgment from a single data source and reduce the interference of cleaning activities on the epitaxial process while ensuring the accuracy of temperature measurement. Attached Figure Description

[0028] Figure 1 This is a structural block diagram of a semiconductor process apparatus in one embodiment; Figure 2 This is a flowchart illustrating a method for treating contamination in a temperature-measuring optical window in one embodiment. Figure 3 This is a structural block diagram of a contamination treatment system for a temperature-measuring optical window in one embodiment; Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0029] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0030] Some exemplary embodiments of this application have been described for illustrative purposes. It should be understood that this application may be implemented in other ways not specifically shown in the accompanying drawings.

[0031] Please see Figure 1 In one exemplary embodiment, this application provides a semiconductor process apparatus, including: a reaction chamber, an optical window assembly, a temperature measurement main optical path assembly, an auxiliary optical path assembly, a purging assembly, a cleaning module, and a contamination treatment system for the temperature measurement optical window.

[0032] A reaction chamber is used to provide an environment for semiconductor epitaxial processes. Exemplarily, the reaction chamber can be circulated with process gases such as a metal-organic source, ammonia, hydrogen, and nitrogen, and epitaxial growth can be performed at a preset temperature and pressure. A temperature measurement channel is provided on the reaction chamber, and an optical window assembly is installed at the temperature measurement channel.

[0033] An optical window assembly includes a temperature-sensing optical window, which is disposed at the temperature-sensing channel of the reaction chamber to provide a channel for optical signal transmission. Exemplarily, the temperature-sensing optical window can be a sapphire window, a calcium fluoride window, or other high-temperature resistant optical windows. The temperature-sensing optical window is mounted on the reaction chamber via a tilted flange to reduce interference from reflected light on the main temperature-sensing optical path.

[0034] The temperature-measuring main optical path component is used to receive radiation signals transmitted through the temperature-measuring optical window within the reaction chamber and generate temperature-measuring main optical path data. Exemplarily, the temperature-measuring main optical path component includes a dual-color infrared temperature measurement component, which is used to acquire radiation intensity signals at a first temperature measurement wavelength and a second temperature measurement wavelength, respectively. The first temperature measurement wavelength is 0.9 μm, and the second temperature measurement wavelength is 1.55 μm. The temperature-measuring main optical path component outputs the first radiation intensity signal, the second radiation intensity signal, and the raw temperature measurement value calculated based on the two radiation intensity signals to the contamination treatment system of the temperature-measuring optical window. This allows the contamination treatment system of the temperature-measuring optical window to calculate a deposition index based on the aforementioned temperature-measuring main optical path data and to make a preliminary judgment on the deposition state of the temperature-measuring optical window based on the deposition index.

[0035] An auxiliary optical path assembly is used to emit an auxiliary monitoring beam and collect auxiliary optical path data obtained after the auxiliary monitoring beam passes through the temperature measuring optical window. Exemplarily, the auxiliary optical path assembly includes an auxiliary monitoring light source, a beam splitter, a reference detector, a transmission detector, a narrowband filter, and a signal amplification circuit. In this embodiment, the auxiliary monitoring light source is a 635nm LED light source. The auxiliary monitoring beam emitted by the auxiliary monitoring light source is collimated and directed towards the temperature measuring optical window. The beam splitter guides a portion of the light output from the auxiliary monitoring light source to the reference detector to form a reference signal; the remaining auxiliary monitoring beam is received by the transmission detector after passing through the temperature measuring optical window to form a transmission signal. The reference signal is used to characterize the output intensity of the auxiliary monitoring light source, and the transmission signal is used to characterize the transmission intensity of the auxiliary monitoring beam after passing through the temperature measuring optical window. The transmission signal and the reference signal are input to the signal amplification circuit, and after amplification, filtering, and analog-to-digital conversion, are input to the contamination treatment system of the temperature measuring optical window to normalize and compensate the transmission signal based on the reference signal. The transmittance data is then calculated based on the normalized and compensated transmission signal, thereby reflecting the change in the transmittance performance of the temperature measuring optical window.

[0036] A purging assembly, disposed on one side of the temperature-sensing optical window, is used to spray inert gas toward the surface of the temperature-sensing optical window to provide a protective gas curtain to the window. Exemplarily, the purging assembly includes a porous spray ring, a mass flow controller, a pressure transmitter, and a check valve.

[0037] A porous spray ring is arranged around the temperature measuring optical window. It has an outer diameter of 70mm, an inner diameter of 45mm, and a thickness of 8mm, and is made of SUS316L. The inner side of the porous spray ring has 48 evenly distributed micro-holes with a diameter of 0.3mm. The spray axis of each micro-hole is inclined downwards at a 45° angle relative to the normal to the surface of the temperature measuring optical window, so that the ejected inert gas forms a stable gas curtain along the surface of the temperature measuring optical window. The distance between the porous spray ring and the window surface is 6mm.

[0038] A mass flow controller is connected to a porous spray ring and is used to regulate the flow rate of inert gas input to the porous spray ring. For example, the mass flow controller is an MFC for controlling the flow rate of nitrogen, with a range of 0-50 slm, a control accuracy of ±1%FS, and a response time less than or equal to 100 ms. The mass flow controller receives a 4-20 mA input control signal, where the 4-20 mA input control signal corresponds to a flow rate adjustment range of 0-50 slm.

[0039] A pressure transmitter is installed on the side wall of the reaction chamber to collect real-time pressure signals within the chamber and output these signals to the contamination treatment system at the temperature-sensing optical window. For example, the pressure transmitter has a range of 0-1000 mbar absolute pressure, an output signal of 4-20 mA, and an accuracy of ±0.25%FS. Based on the reaction chamber pressure, pressure change rate, and process stage information collected by the pressure transmitter, the contamination treatment system at the temperature-sensing optical window generates a target flow rate setpoint for the mass flow controller, thereby dynamically adjusting the purge flow rate of the protective gas curtain to match the current process conditions.

[0040] A check valve is installed between the outlet of the mass flow controller and the porous spray ring to prevent process gas, reaction byproducts, or particulate matter in the reaction chamber from flowing back into the purge line during gas outages or pressure fluctuations. For example, the check valve is a spring-loaded one-way valve with an opening pressure of 0.5 bar. By installing a check valve, the safety and stability of the purge assembly under pressure fluctuation conditions can be improved.

[0041] A cleaning module is used to perform in-situ cleaning of a temperature-sensing optical window based on a contamination treatment strategy. Exemplarily, the cleaning module includes a pulsed laser, a two-dimensional scanning galvanometer, a field mirror, and a laser drive power supply. The pulsed laser outputs laser light of at least one cleaning wavelength, such as a 355nm or 1064nm laser. The two-dimensional scanning galvanometer controls the laser beam to move along a preset scanning path on the surface of the temperature-sensing optical window. The contamination treatment system can determine laser cleaning parameters, such as laser wavelength, pulse energy, scanning path, laser dwell time, or scanning speed, based on the deposition type, deposition thickness, or deposition optical parameters, and control the cleaning module to perform in-situ cleaning of the temperature-sensing optical window.

[0042] The contamination treatment system for the temperature measurement optical window is used to determine the contamination treatment strategy based on the main optical path data, auxiliary optical path data, and process operation data.

[0043] Specifically, the semiconductor process equipment of this application also includes a process control system. The process operation data is provided by the process control system and is used to characterize the current operating status of the semiconductor process equipment. For example, the process operation data includes at least one of process stage information, reaction chamber pressure, pressure change rate, time since last cleaning, and cleaning history data.

[0044] During equipment operation, the contamination treatment system of the temperature-measuring optical window acquires temperature-measuring main optical path data, auxiliary optical path data, and process operation data. The process operation data includes process stage information. Based on the temperature-measuring main optical path data, a deposition index is calculated, and based on the deposition index, the main optical path estimation result is determined. The main optical path estimation result includes thickness estimation results and type prediction results. Based on the auxiliary optical path data, transmittance data is calculated. When at least one of the deposition index and transmittance data meets the preset inversion conditions, multi-parameter Monte Carlo inversion is performed based on the temperature-measuring main optical path data and auxiliary optical path data to obtain the inversion result. The inversion result includes the inverted deposition thickness and the inverted deposition optical parameters. Based on the main optical path estimation result, transmittance data, and inversion result, multi-source integrated processing is performed to determine the deposition state information of the temperature-measuring optical window. The deposition state information includes at least one of fused deposition thickness, fused deposition optical parameters, and deposition type. Based on the deposition state information and process stage information, a contamination treatment strategy is determined. The contamination treatment strategy includes at least one of temperature compensation strategy and in-situ cleaning strategy. The aforementioned semiconductor process equipment, by setting a temperature-measuring optical window at the temperature measurement channel of the reaction chamber and configuring a main temperature measurement optical path component and an auxiliary temperature measurement optical path component, enables the equipment to simultaneously acquire temperature measurement main optical path data and auxiliary optical path data, thereby enabling online monitoring of the deposition or contamination status of the temperature-measuring optical window without affecting the normal operation of the epitaxial process.

[0045] Furthermore, this application collects radiation signals within the reaction cavity transmitted through the temperature-measuring optical window using the temperature-measuring main optical path component. This allows for the reuse of existing temperature-measuring signals to achieve preliminary identification of the window deposition state, reducing additional hardware modifications and improving the device's real-time sensing capability for changes in window contamination.

[0046] Furthermore, this application emits an auxiliary monitoring beam and collects transmission and reference signals through an auxiliary optical path component, enabling independent detection of the transmittance performance of the temperature measuring optical window. The reference signal also reduces the impact of auxiliary light source power fluctuations on the detection results, thereby improving the reliability of the window transmittance data.

[0047] Furthermore, by setting up a purging assembly and dynamically adjusting the target purging flow rate according to the reaction chamber pressure, pressure change rate, and process stage information, this application can form a protective gas curtain on the surface of the temperature measuring optical window that matches the current process state, thereby reducing the probability of precursors, by-products, or particulate matter depositing on the window surface.

[0048] Furthermore, this application sets up a cleaning module and controls a pulsed laser, a two-dimensional scanning galvanometer, and a field lens to perform in-situ cleaning of the temperature measuring optical window according to the contamination treatment strategy. This enables directional cleaning when the window deposition reaches the cleaning conditions, reducing the frequency of downtime for disassembly and cleaning or manual maintenance.

[0049] Furthermore, this application integrates the main optical path data, auxiliary optical path data, and process operation data of temperature measurement through a contamination treatment system for the temperature measurement optical window. It can determine the temperature compensation strategy or in-situ cleaning strategy based on the window deposition state, thereby correcting the temperature measurement results before the window is cleaned and performing cleaning at a suitable process stage to avoid interference with the epitaxial growth process.

[0050] In summary, the semiconductor process equipment of this application embodiment can achieve coordinated control of online monitoring of the contamination status of the temperature measurement optical window, dynamic purging protection, temperature compensation, and in-situ cleaning, thereby improving the accuracy and stability of the temperature measurement results of the epitaxial process, reducing the impact of window contamination on process control, and enhancing the continuous operation capability of the semiconductor process equipment.

[0051] Please see Figure 2 In one exemplary embodiment, this application provides a method for treating contamination in a temperature-measuring optical window. This method is applicable to the semiconductor process equipment provided in the above embodiments and specifically includes the following steps: Step 202: Acquire the main optical path data, auxiliary optical path data, and process operation data for temperature measurement; the process operation data includes process stage information.

[0052] Specifically, the temperature measurement main optical path data and auxiliary optical path data have been preprocessed. For example, the raw data acquired based on the temperature measurement main optical path component undergoes pre-amplification, noise filtering, dark current subtraction, analog-to-digital conversion, and sample averaging to obtain the temperature measurement main optical path data, thereby reducing the impact of detector noise, electronic noise, and transient disturbances on subsequent deposition index calculations. The raw data acquired based on the auxiliary optical path component undergoes signal amplification, bandpass filtering, background subtraction, analog-to-digital conversion, and outlier removal, and is further normalized in the contamination processing system of the temperature measurement optical window to obtain the auxiliary optical path data, thereby mitigating the impact of auxiliary monitoring light source power fluctuations, light source aging, or temperature drift on the transmission signal.

[0053] Step 204: Calculate the deposition index based on the temperature measurement main optical path data, and determine the main optical path estimation result based on the deposition index; the main optical path estimation result includes the thickness estimation result and the type prediction result.

[0054] Specifically, the deposition index is used to characterize the degree of influence of deposits on the surface of the temperature measurement optical window on the signal of the main temperature measurement optical path.

[0055] When no significant deposits occur on the surface of the temperature-measuring optical window, the signal relationship corresponding to different temperature-measuring wavelengths in the main temperature-measuring optical path is usually maintained near the clean baseline; whereby the clean baseline refers to the reference correspondence between the data of the main temperature-measuring optical path and the temperature when the temperature-measuring optical window is in a clean state.

[0056] When deposits appear on the surface of the temperature-measuring optical window, the radiation intensity signals corresponding to different temperature-measuring wavelengths will attenuate or shift to varying degrees, causing the temperature-measuring main optical path signal to deviate from the clean baseline. Therefore, the contamination treatment system can calculate the deposition index based on the temperature-measuring main optical path data and use the deposition index to reflect the current deposition state of the temperature-measuring optical window.

[0057] The main optical path estimation result refers to the preliminary estimation result obtained based on the temperature measurement main optical path data. Among them, the thickness estimation result is used to characterize the estimated thickness of the deposition layer on the surface of the temperature measurement optical window; the type prediction result is used to characterize the preliminary type of the deposit. For example, the deposition type includes absorptive deposition, transparent thin film deposition and complex deposition. Step 206: Calculate the transmittance data based on the auxiliary optical path data.

[0058] Specifically, transmittance data is used to characterize the degree to which the light intensity of the auxiliary monitoring beam is maintained after passing through the temperature measuring optical window, and also to reflect the current light transmission performance of the temperature measuring optical window. When the temperature measuring optical window is clean, the transmission intensity of the auxiliary monitoring beam after passing through the temperature measuring optical window remains within a preset reference range; when there are deposits or contaminants on the surface of the temperature measuring optical window, the auxiliary monitoring beam will experience absorption, scattering, or reflection losses when passing through the temperature measuring optical window, causing a decrease in transmission intensity. Therefore, the degree of contamination or deposits on the temperature measuring optical window can be determined through transmittance data. Step 208: If at least one of the deposition index and transmittance data meets the preset inversion conditions, perform multi-parameter Monte Carlo inversion based on the temperature measurement main optical path data and auxiliary optical path data to obtain the inversion results; the inversion results include the inverted deposition thickness and the inverted deposition optical parameters.

[0059] Specifically, when the deposition state on the surface of the temperature-measuring optical window is relatively simple, such as when the deposition index changes monotonically and the transmittance data is consistent with the estimation result of the main optical path, the deposition state can be determined based on the temperature-measuring main optical path data and the auxiliary optical path data. For such scenarios, there is no need to initiate multi-parameter Monte Carlo inversion, thus reducing computational resource consumption.

[0060] In actual epitaxial processes, the deposits on the surface of the temperature-measuring optical window do not always exhibit simple absorption and attenuation characteristics. For example, if a transparent thin film or multilayer composite deposition forms on the window surface, the signal in the main temperature-measuring optical path may oscillate; if both absorbing contaminants and a transparent thin film are present on the window surface, inconsistencies may arise between the deposition index and transmittance data; and under the combined effects of cavity thermal radiation, thin film interference, scattering, and differences in absorption at different wavelengths, it is difficult to distinguish between changes in deposition thickness and changes in deposition optical parameters based solely on the deposition index or single transmittance data. In these scenarios, if simple thresholds or empirical models are still used, it is easy to misjudge the transparent thin film as absorbing contaminants, or to mistake changes in optical parameters for an increase in deposition thickness, leading to incorrect temperature compensation or inappropriate selection of cleaning parameters.

[0061] Based on this, this application initiates multi-parameter Monte Carlo inversion when at least one of the deposition index and transmittance data meets a preset inversion condition. The multi-parameter Monte Carlo inversion uses deposition thickness and deposition optical parameters as the parameters to be inverted, and the deposition optical parameters include at least one of refractive index and extinction coefficient. Measured transmittance data at multiple wavelengths are constructed based on the temperature-measuring main optical path data and auxiliary optical path data. Then, theoretical transmittance data corresponding to different candidate parameters are calculated through Monte Carlo photon transmission simulation, and the candidate parameters are updated based on the degree of matching between the theoretical transmittance data and the measured transmittance data. Finally, the inverted deposition thickness and inverted deposition optical parameters are obtained.

[0062] Step 210: Perform multi-source integrated processing based on the main optical path estimation results, transmittance data and inversion results to determine the deposition state information of the thermometric optical window; wherein, the deposition state information includes at least one of the following: fused deposition thickness, fused deposition optical parameters and deposition type.

[0063] Specifically, this application employs a tiered integrated processing approach based on the main optical path estimation results, transmittance data, and the reliability of the inversion results. For conventional scenarios where the deposition index changes steadily and the transmittance data aligns with the main optical path estimation results, the depositional state information is primarily determined based on the main optical path estimation results and transmittance data, eliminating the need to introduce inversion results and thus reducing computational resource consumption. For complex scenarios where the deposition index exhibits oscillating changes, the transmittance data is inconsistent with the main optical path estimation results, the model uncertainty is high, or multi-parameter Monte Carlo inversion has been triggered, the inversion results are used as one of the data sources for multi-source integrated processing to improve the accuracy of determining depositional thickness and depositional optical parameters.

[0064] Step 212: Determine the contamination treatment strategy based on the deposition state information and process stage information; the contamination treatment strategy includes at least one of the temperature compensation strategy and the in-situ cleaning strategy.

[0065] Specifically, the contamination treatment strategy is used to indicate the treatment methods that can be performed for the current deposition state of the temperature measuring optical window. Among them, the temperature compensation strategy is used to correct the original temperature measurement value output by the main temperature measuring optical path when there is deposition on the temperature measuring optical window but cleaning has not yet been performed; the in-situ cleaning strategy is used to perform cleaning on the temperature measuring optical window when the deposition degree reaches the cleaning condition and the current process stage allows cleaning.

[0066] The aforementioned method for handling contamination of the temperature-measuring optical window involves acquiring data from the main temperature-measuring optical path, auxiliary optical path, and process operation data. Based on the main temperature-measuring optical path data, a deposition index is calculated, and the estimation result of the main optical path is determined. The original main temperature-measuring optical path is reused to achieve online preliminary identification of the window's deposition state. Then, transmittance data is calculated based on the auxiliary optical path data to supplement and characterize the actual transmittance performance of the temperature-measuring optical window, improving the reliability of the deposition degree judgment. Multi-parameter Monte Carlo inversion is performed when preset inversion conditions are met, obtaining more accurate deposition thickness and deposition optical parameters in complex deposition or when conventional judgments are uncertain. Finally, by comprehensively processing the main optical path estimation results, transmittance data, and inversion results from multiple sources, the deposition state information of the temperature-measuring optical window is determined. Based on the deposition state information and process stage information from the process operation data, a contamination handling strategy is determined. This method reduces the risk of misjudgment from a single data source and minimizes the interference of cleaning activities on the epitaxial process while ensuring the accuracy of temperature measurement.

[0067] Optionally, when the temperature measurement main optical path data includes a first radiation intensity signal corresponding to a first temperature measurement wavelength and a second radiation intensity signal corresponding to a second temperature measurement wavelength, the deposition index is calculated based on the temperature measurement main optical path data, including: The original temperature value is calculated based on the first and second radiation intensity signals. The original intensity ratio at the current moment is calculated based on the first and second radiation intensity signals. The corresponding clean intensity ratio is determined from the pre-established clean baseline relationship based on the original temperature value. The clean baseline relationship is used to characterize the correspondence between different temperature values ​​and the corresponding clean intensity ratio under clean temperature measurement optical window conditions. The relative deviation information is determined based on the original intensity ratio and the clean intensity ratio. The deposition index is obtained by filtering the relative deviation information.

[0068] Specifically, obtain the current sampling time The first radiation intensity signal corresponding to the first temperature measurement wavelength The second radiation intensity signal corresponding to the second temperature measurement wavelength ,based on and Perform dual-color temperature measurement calculations to obtain the original temperature values. The original temperature measurement value The temperature value before window contamination compensation is expressed as:

[0069] Clean baseline relationship This is pre-established under conditions where the temperature-measuring optical window is clean and the reaction chamber is in a calibrated allowable state, used to characterize the correspondence between different temperature values ​​and the cleanliness intensity ratio. For example, the reaction chamber temperature is sequentially set to eight calibration points: 700℃, 800℃, 900℃, 1000℃, 1050℃, 1100℃, 1150℃, and 1200℃, with each temperature point held for 30 seconds after stabilization. At each temperature point, 200 sets of data are continuously collected at a sampling rate of 100Hz. , ), calculate the ratio of each group The average of the ratios is taken to obtain the corresponding temperature for that group. .right( Perform cubic spline interpolation to obtain the clean baseline relationship. .

[0070] Furthermore, based on the original temperature measurement values From the clean baseline relationship The corresponding cleanliness intensity ratio can be obtained by querying or interpolating. Its expression is: .

[0071] Subsequently, based on the original strength ratio Ratio of cleanliness intensity Calculate relative deviation information Its expression is:

[0072] Relative deviation information This is used to characterize the deviation of the current temperature measurement main optical path signal from the clean window state. When deposits form on the surface of the temperature measurement optical window, the attenuation of radiation intensity at different temperature measurement wavelengths varies, resulting in a difference in the original intensity ratio. relative to cleanliness intensity ratio An offset is generated, which reflects the effect of window deposition on the main optical path for temperature measurement.

[0073] To reduce the impact of transient noise, thermal radiation disturbances, and electronic noise on deposition judgment, this application also includes information on relative deviation. Low-pass filtering is performed to obtain the filtered deposition index. In some embodiments, the filtering process employs a first-order low-pass filter, the expression of which is:

[0074] in, The deposition index at the current moment. The deposition index is the value at the previous sampling time. These are the filter coefficients. Further, the filter coefficients... Determine using the following expression:

[0075] in, The cutoff frequency, The sampling period is denoted as . For example, the cutoff frequency in this embodiment is . The sampling period is ,but Approximately .

[0076] Optionally, the main optical path estimation result is determined based on the deposition index, including: Based on the deposition index at different times, the variation characteristics of the deposition index are determined; based on the deposition index and the variation characteristics of the deposition index, the type prediction result is determined; wherein, the type prediction result includes at least one of absorptive deposition, transparent film deposition, and complex deposition; if the type prediction result is absorptive deposition, the thickness estimation result is determined based on the deposition index and the pre-established empirical model of deposition thickness; the type prediction result and / or the thickness estimation result are determined as the main optical path estimation result.

[0077] Specifically, the variation characteristics include at least one of the following: sedimentation index change rate, monotonic change trend of sedimentation index, and sedimentation index oscillation characteristics. For example, in this embodiment, the sedimentation index change rate is used, calculated using the central difference method with a window width of 10 sampling points. The expression is as follows:

[0078] in, For the current moment The sedimentation index, This represents the sampling time interval.

[0079] Furthermore, this application is based on the sedimentation index and its rate of change This allows for the prediction of the deposition type on the surface of the temperature-measuring optical window. For example, when... ,and When this occurs, it indicates that the signal in the main optical path for temperature measurement continuously decays over time, and the predicted type is absorbing deposition, such as carbonaceous deposition. ,and When the temperature measurement signal is close to the clean baseline or only slightly affected by the transparent film layer, the type prediction result is transparent thin film deposition or a clean state. When the deposition index shows alternating positive and negative changes or non-monotonic oscillations, it indicates that the deposition layer on the window surface may produce thin film interference effects or complex optical effects, and the type prediction result is complex deposition.

[0080] Furthermore, when the predicted deposition type is absorptive deposition, this application determines the thickness estimate based on the deposition index and a pre-established empirical model of deposition thickness. For example, the empirical model expression corresponding to absorptive deposition is:

[0081] in, The sedimentation index, For the deposition thickness, and These are the model coefficients obtained through pre-calibration. For example, , Then, the thickness estimation result is obtained by inverse solving the empirical model:

[0082] in, The thickness estimation result is based on the temperature measurement main optical path.

[0083] Using the above scheme, the changing characteristics of the deposition index can identify the dynamic impact of window deposition on the temperature measurement main optical path signal. Combining the deposition index and its rate of change to determine the type prediction result allows for preliminary differentiation of different deposition types, providing a basis for subsequent compensation model selection and cleaning parameter selection. When the type prediction result is absorptive deposition, using an empirical model to calculate the thickness estimation result can also quickly obtain a preliminary estimate of the deposition thickness without immediately initiating a complex inversion, reducing computational resource consumption. Using the type prediction result and thickness estimation result as the main optical path estimation result allows for full reuse of existing data from the temperature measurement main optical path, enabling online preliminary identification of the deposition state of the temperature measurement optical window, providing basic data for subsequent multi-source integrated processing.

[0084] Optionally, when the auxiliary optical path data includes the transmission signal acquired after the auxiliary monitoring beam passes through the temperature-measuring optical window, and the reference signal used to characterize the output intensity of the auxiliary monitoring light source, the transmittance data is calculated based on the auxiliary optical path data, including: The transmitted signal and reference signal are preprocessed to obtain the transmitted signal value and reference signal value; the baseline transmitted signal value and baseline reference signal value under clean conditions are obtained; based on the transmitted signal value, reference signal value, baseline transmitted signal value, and baseline reference signal value, the transmittance data is calculated, and its expression is:

[0085] in, For transmittance data, This is the transmitted signal value. As a reference signal value, As the reference transmission signal value, This is the reference signal value.

[0086] Specifically, the preprocessing process includes at least one of the following: averaging multiple samples, background subtraction, filtering, and outlier removal.

[0087] Furthermore, this application pre-records the reference transmission signal value when the temperature measuring optical window is in a clean state. and reference signal value Among them, the reference transmission signal value The reference signal value characterizes the transmission intensity of the auxiliary monitoring beam after passing through the temperature measuring optical window under clean window conditions. The output intensity of the auxiliary monitoring light source is used to characterize the cleanliness calibration.

[0088] In the current detection process, transmittance data is calculated according to the above expression, where, Characterizes the degree of attenuation of the current transmission intensity relative to the clean window state. Used to normalize and compensate for changes in the output intensity of auxiliary monitoring light sources.

[0089] Using the above scheme, preprocessing can reduce the impact of background light, transient noise, and abnormal sampling points on transmittance calculation; introducing a reference transmission signal value and a reference signal value can compare the current detection results with the clean window status, thereby quantitatively characterizing the transmittance performance changes of the temperature measuring optical window; using the reference signal to normalize and compensate the transmission signal can weaken the impact of auxiliary monitoring light source power fluctuations, light source aging, or temperature drift on transmittance data, and improve the reliability of transmittance data.

[0090] Optionally, the determination method for satisfying the preset inversion conditions includes: If at least one of the following conditions is met: the magnitude of change of the deposition index is greater than a preset magnitude threshold, the deposition index oscillates a preset number of times within a preset time period, the model uncertainty corresponding to the deposition index is greater than a preset uncertainty threshold, or the deviation between the transmittance data and the transmittance predicted based on the temperature measurement main optical path data is greater than a preset deviation threshold, then the preset inversion conditions are considered to be met.

[0091] For example, the model uncertainty in this embodiment Determine using the following expression:

[0092] in, The standard deviation of the most recently preset number of deposition indices. For transmittance data obtained from auxiliary optical path data, The transmittance is predicted based on the temperature measurement main optical path data. Let V be the variance of the rate of change of the sedimentation index. , , These are the weighting coefficients.

[0093] Furthermore, the transmittance predicted based on the temperature measurement main optical path data is calculated using a pre-trained random forest model. In this embodiment, the random forest model uses 100 trees with a maximum depth of 10.

[0094] Optionally, if at least one of the deposition index and transmittance data meets a preset inversion condition, a multi-parameter Monte Carlo inversion is performed based on the temperature measurement main optical path data and auxiliary optical path data to obtain the inversion results, including: Under the premise of meeting the preset inversion conditions, the first transmittance data at at least one temperature measurement wavelength is determined based on the main temperature measurement optical path data, and the second transmittance data at an auxiliary wavelength is determined based on the auxiliary optical path data. After combination, the measured transmittance data at multiple wavelengths are obtained. Using the deposition thickness and deposition optical parameters as the parameters to be inverted, a multi-parameter inversion model is constructed to characterize the deposition layer on the surface of the temperature measurement optical window. The deposition optical parameters include refractive index and extinction coefficient. Monte Carlo photon transmission simulation is performed based on the multi-parameter inversion model to obtain the theoretical transmittance data corresponding to different parameters to be inverted. According to the degree of matching between the measured transmittance data and the theoretical transmittance data, a likelihood function is constructed, and Bayesian sampling is performed on the parameters to be inverted based on the likelihood function to obtain the inversion results.

[0095] For example, the above-mentioned multiple wavelengths include a first temperature measurement wavelength, a second temperature measurement wavelength, and an auxiliary monitoring wavelength, and the measured transmittance data is denoted as: ;in, For the first One wavelength, The number of wavelengths involved in the inversion.

[0096] Furthermore, a multi-parameter inversion model is constructed using deposition thickness and deposition optical parameters as the parameters to be inverted. The vector of parameters to be inverted is denoted as: ;in, For the deposition thickness, For refractive index, Extinction coefficient, For surface roughness. In some embodiments, if surface roughness is not considered, the parameter vector to be inverted can be further simplified to: .

[0097] In the Monte Carlo photon transport simulation, for each set of candidate parameters This simulates the absorption, scattering, reflection, and transmission processes of photons within the deposition layer and the thermometric optical window. For the absorption process in the deposition layer, the absorption probability is determined by the following expression:

[0098]

[0099] in, For absorption probability, The absorption coefficient is... For photon propagation step size, Where is the wavelength. After the simulation, the theoretical transmittance data is obtained based on the ratio between the final transmitted photon count and the emitted photon count, and its expression is:

[0100] in, This represents the number of transmitted photons. The number of photons emitted.

[0101] Furthermore, based on measured transmittance data Compared with theoretical transmittance data The likelihood function is constructed based on the degree of matching between the wavelengths. For example, when the measurement errors at each wavelength follow a Gaussian distribution, the likelihood function is expressed as:

[0102] in, For the first The standard deviation of the measurement error corresponding to each wavelength This represents the total number of wavelengths.

[0103] Optionally, Bayesian sampling is performed on the parameters to be inverted based on the likelihood function to obtain the inversion results, including: Set a prior distribution for the parameters to be inverted; generate a candidate parameter set based on the prior distribution; determine the acceptance probability of the candidate parameter set based on the difference between the theoretical transmittance data and the measured transmittance data corresponding to the candidate parameter set; iteratively sample the candidate parameter set according to the acceptance probability; determine the inversion result based on the sampling results after the convergence condition is met.

[0104] Specifically, Bayesian sampling is performed on the parameters to be inverted based on the likelihood function. First, a priori distributions are set for the parameters to be inverted; for example, the deposition thickness follows a uniform distribution within a preset thickness range, the refractive index follows a truncated normal distribution, and the extinction coefficient follows a log-normal distribution. Then, candidate parameter sets are generated based on the prior distributions. And calculate the theoretical transmittance data corresponding to the candidate parameter groups.

[0105] In some embodiments, the Metropolis-Hastings sampling method is used to determine the acceptance probability of the candidate parameter group, and the expression for the acceptance probability is:

[0106] in, For the current parameter group, For candidate parameter groups, and These are the likelihood function values ​​for the current parameter set and the candidate parameter set, respectively. and These are the prior probabilities corresponding to the current parameter set and the candidate parameter set, respectively.

[0107] When a candidate parameter set is accepted, it is used as the current parameter set for the next iteration; when a candidate parameter set is not accepted, the current parameter set remains unchanged. After multiple rounds of iterative sampling, if the sampling chain meets the convergence condition, the inversion result is determined based on the converged sampling results. The inversion result includes at least one of the inverted deposition thickness, inverted refractive index, and inverted extinction coefficient. For example, meeting the convergence condition includes the change in the mean parameter being less than a preset proportion or the convergence statistic being less than a preset threshold. The above scheme only initiates multi-parameter Monte Carlo inversion when the preset rebound conditions are met, reducing computational overhead in conventional scenarios. Using deposition thickness and deposition optical parameters as inversion parameters allows for the simultaneous analysis of the geometric thickness and optical properties of the deposition layer, avoiding misinterpretation of optical parameter changes as thickness changes. Theoretical transmittance data obtained through Monte Carlo photon transport simulation can describe the absorption, scattering, reflection, and transmission processes in the deposition layer, improving the matching degree between the theoretical model and the actual optical path state in complex deposition scenarios. Bayesian sampling based on the likelihood function can iteratively search for parameter combinations matching the measured transmittance data in the parameter space, obtaining more reliable inversion deposition thickness and inversion deposition optical parameters. Ultimately, in scenarios where transparent films, absorptive depositions, and complex interferometric depositions are difficult to judge using empirical models, the accuracy of deposition state identification through the thermometric optical window is improved, providing an accurate basis for subsequent multi-source integrated processing, temperature compensation, and in-situ cleaning parameter selection.

[0108] Optionally, based on the main optical path estimation results, transmittance data, and inversion results, multi-source integrated processing is performed to determine the deposition state information of the thermometric optical window, including: Based on transmittance data, the auxiliary optical path estimation results are determined; the auxiliary optical path estimation results include auxiliary deposition thickness and auxiliary deposition optical parameters; based on the main optical path estimation results, auxiliary optical path estimation results, and inversion results, thickness fusion processing is performed to obtain the fused deposition thickness; based on transmittance data and inversion results, the fused deposition optical parameters are determined; based on the variation characteristics of the fused deposition optical parameters and deposition index, the type prediction results are corrected to obtain the deposition type; the variation characteristics are determined based on the deposition index at different times.

[0109] Specifically, the auxiliary deposition thickness is used to characterize the deposition layer thickness derived from the transmittance data, and the auxiliary deposition optical parameters are used to characterize at least one of the refractive index and extinction coefficient derived from the transmittance data and the thin film optical model.

[0110] For example, if the type prediction result is absorptive deposition, then based on the transmittance data... Thickness estimation results obtained from main optical path estimation and auxiliary monitoring wavelength The auxiliary extinction coefficient is derived by deduction. The expression is:

[0111] in, For transmittance data, For the thickness estimation results, To assist in monitoring wavelength, To assist in the extinction coefficient.

[0112] If the predicted type is transparent thin film deposition, then parameter inversion is performed based on transmittance data and a preset thin film optical model. The theoretical transmittance output by the preset thin film optical model is... Its expression is:

[0113]

[0114] in, The refractive index of air, The refractive index of the window substrate, The refractive index of the deposited layer, For the thickness of the sedimentary layer, To assist in monitoring the wavelength, the auxiliary deposition thickness and auxiliary optical parameters are obtained by solving the following optimization problem: ;in, The transmittance data is used to obtain the auxiliary optical path estimation result.

[0115] Optionally, based on the main optical path estimation results, auxiliary optical path estimation results, and inversion results, thickness fusion processing is performed to obtain the fused deposition thickness, including: Based on the estimation results of the main optical path, the estimation results of the auxiliary optical path, and the reliability of the inversion results, their respective fusion weights are determined. In the case where there are no inversion results or the reliability of the inversion results does not meet the preset reliability conditions, the fusion weight corresponding to the inversion results is set to zero. The fusion weights are used to perform thickness fusion processing to obtain the fused deposition thickness.

[0116] Specifically, the thickness estimation result in the main optical path estimation result is denoted as... The auxiliary deposition thickness in the auxiliary optical path estimation result is denoted as The inversion sediment thickness in the inversion results is denoted as .

[0117] If the inversion result exists and its confidence level meets a preset confidence level condition, the inversion result participates in the thickness fusion process. For example, if the half-width of the confidence interval of the inversion result is less than a preset width threshold, the inversion result is determined to meet the preset confidence level condition. At this time, the fused deposition thickness... Determine using the following expression: .

[0118] If multi-parameter Monte Carlo inversion is not performed, or the confidence level of the inversion result does not meet the preset confidence level condition, but the auxiliary optical path estimation result is valid, the fusion weight corresponding to the inversion result is reset to zero. In this case, the fused deposition thickness is... Determine using the following expression: .

[0119] If both the auxiliary optical path estimation results and the inversion results are unavailable, the thickness estimation results from the main optical path estimation results will be used. As the thickness of fusion deposition .

[0120] It should be understood that the above fusion weight values ​​are only examples, and can be set according to needs in actual applications.

[0121] Optionally, based on transmittance data and inversion results, the optical parameters for fusion deposition are determined, including: Based on transmittance data and a preset thin film optical model, parameter inversion is performed to determine auxiliary optical parameters that match the theoretical transmittance with the transmittance data. If there are no inversion results or the inversion results do not meet the preset confidence level, the auxiliary optical parameters are determined as fusion deposition optical parameters. If the inversion results meet the preset confidence level, the inverted optical parameters in the inversion results are determined as fusion deposition optical parameters.

[0122] Specifically, the inversion optical parameters include the inversion refractive index. and inversion extinction coefficient The fused deposition optical parameters include the refractive index and extinction coefficient, which are ultimately used to determine the deposition type.

[0123] Furthermore, based on the variation characteristics of the fusion deposition optical parameters and deposition index, the type prediction result is corrected to obtain the deposition type. For example, when the extinction coefficient in the fusion deposition optical parameters is greater than the first type threshold, the deposition type is confirmed or corrected to be an absorptive deposition; when the extinction coefficient is less than the second type threshold, the deposition type is confirmed or corrected to be a transparent thin film deposition; when the extinction coefficient is in the intermediate range, or the deposition index shows oscillating changes within a preset time, the deposition type is confirmed or corrected to be a mixed deposition or complex deposition. For example, the first type threshold is set to 0.1, and the second type threshold is set to 0.01.

[0124] By employing the above scheme, this application first determines the auxiliary optical path estimation result based on transmittance data. This allows the transmittance variation at the auxiliary wavelength to supplement the main optical path estimation result, improving the reliability of deposition thickness and optical parameter judgment. The thickness is then fused based on the reliability of the signal source, combining the main optical path estimation result, the auxiliary optical path estimation result, and the inversion result. This allows for different weights to be applied in deposition scenarios of varying complexity, preventing low-reliability data from affecting the fused deposition thickness. Furthermore, this application prioritizes the use of inversion optical parameters when the inversion result is reliable, and uses auxiliary optical parameters when the inversion result is unreliable or nonexistent. This approach balances inversion accuracy in complex scenarios with computational efficiency in conventional scenarios. Moreover, by correcting the type prediction result based on the fused deposition optical parameters and deposition index variation characteristics, the risk of type misjudgment caused by relying solely on the main optical path prediction is reduced, improving the accuracy of deposition type identification.

[0125] Optionally, based on sedimentation state information and process stage information, a contamination treatment strategy is determined, including: Based on the fusion deposition thickness and deposition type, the deposition degree of the temperature-measuring optical window is determined; if the deposition degree reaches the temperature compensation threshold but does not reach the cleaning threshold, a contamination treatment strategy including a temperature compensation strategy is determined; if the deposition degree reaches the cleaning threshold and the process stage information indicates that the process is currently in a cleanable stage, a contamination treatment strategy including an in-situ cleaning strategy is determined; if the deposition degree reaches the cleaning threshold and the process stage information indicates that the process is not currently in a cleanable stage, a contamination treatment strategy including a temperature compensation strategy is determined.

[0126] Specifically, after obtaining the deposition state information, the deposition degree of the temperature-measuring optical window is determined based on the fused deposition thickness and deposition type. The deposition degree characterizes the extent of the impact of the current deposition layer on the main temperature-measuring optical path. This application considers both the fused deposition thickness and deposition type in determining the deposition degree because the mechanisms by which absorptive deposition and transparent film deposition affect the main temperature-measuring optical path differ. Absorptive deposition primarily causes radiation signal attenuation, while transparent film deposition primarily causes changes in reflectivity or interference effects.

[0127] If the deposition level reaches the temperature compensation threshold but not the cleaning threshold, it indicates that the current deposition has affected the original temperature measurement value, but has not yet reached the level requiring cleaning. In this case, the contamination treatment strategy includes a temperature compensation strategy. If the deposition level reaches the cleaning threshold, and the process stage information indicates that the current process is in a cleanable stage such as standby, material change, or cooling, the contamination treatment strategy includes an in-situ cleaning strategy. If the deposition level reaches the cleaning threshold, but the process stage information indicates that the current process is in an unsuitable cleaning stage such as the growth stage, the contamination treatment strategy includes a temperature compensation strategy, and the in-situ cleaning strategy is postponed. By adopting the above scheme, this application can avoid the critical growth stage during cleaning, reducing disturbance to the epitaxial growth environment.

[0128] For example, for absorptive deposition, the fusion deposition thickness corresponding to the cleaning threshold is the first cleaning thickness threshold; for transparent film deposition, the fusion deposition thickness corresponding to the cleaning threshold is the second cleaning thickness threshold. Since different deposition types have different degrees of influence on the temperature measurement signal, the first and second cleaning thickness thresholds are taken as different values. Specifically, in this embodiment, the first cleaning thickness threshold is 50 nm, and the second cleaning thickness threshold is 150 nm. In more severe cases, such as when absorptive deposition reaches 120 nm or transparent film deposition reaches 300 nm, an emergency cleaning alarm signal is immediately output even during the growth stage, so that engineers can handle it promptly.

[0129] Optionally, the temperature compensation strategy includes: determining the temperature compensation amount based on the mapping model between depositional state information and temperature deviation; correcting the original temperature measurement value output by the main optical path of temperature measurement based on the temperature compensation amount to obtain the corrected temperature measurement value; wherein, the mapping model is a nonlinear model established based on calibration data, and the nonlinear model is used to characterize the temperature measurement deviation corresponding to different depositional state information; the nonlinear model includes at least one of a polynomial model, a lookup table model, or a piecewise fitting model.

[0130] Specifically, the mapping model is used to characterize the temperature measurement deviation corresponding to different sedimentation state information, including at least one of the polynomial model, lookup table model or piecewise fitting model.

[0131] For example, for absorptive deposition, this embodiment corrects the radiation intensity in the main thermometric optical path based on the Beer-Lambert law. For wavelength... Permeability factor of sediment layer The expression is:

[0132] in, To integrate the deposition thickness, The absorption coefficient is determined by the following expression:

[0133] in, The extinction coefficient is used. The temperature compensation is obtained by correcting the intensity ratio of the two wavelengths in the main temperature measurement optical path based on the transmission factor of the deposition layer. The temperature compensation expression is:

[0134] in, and These are the first and second temperature measurement wavelengths, respectively. and These are the transmission factors of the deposition layer at two different temperature measurement wavelengths. is the radiation constant.

[0135] For example, for transparent thin film deposition, the temperature compensation amount is determined based on the change in thin film reflectivity, and its compensation expression is as follows:

[0136] in, and The thin film reflectance at two different temperature measurement wavelengths are shown. This is the calibration coefficient. In actual operation, the fused deposition thickness, deposition type, and fused deposition optical parameters can be input into a lookup table model, and the temperature compensation amount can be obtained through interpolation. Subsequently, the original temperature measurement value output from the main temperature measurement optical path is corrected based on the temperature compensation amount to obtain the corrected temperature measurement value: .in, The original temperature measurement value. This is the temperature compensation amount. This is the corrected temperature measurement value.

[0137] Optionally, the in-situ cleaning strategy includes: performing multi-point scanning on the temperature-measuring optical window to generate a contamination distribution map; determining the laser scanning path and the laser dwell time and / or scanning speed corresponding to each scanning area based on the contamination distribution map, wherein areas with a contamination level greater than a preset threshold are designated as the first area, areas less than or equal to the preset threshold are designated as the second area, the laser dwell time corresponding to the first area is greater than the laser dwell time corresponding to the second area, and / or the scanning speed corresponding to the first area is less than the scanning speed corresponding to the second area; acquiring auxiliary optical path data in real time during the cleaning process, and determining the real-time transmittance data of the current area based on the auxiliary optical path data; stopping or skipping subsequent cleaning of the current area when the real-time transmittance data recovers to the target transmittance.

[0138] Specifically, this application differentiates the laser scanning energy according to the contamination distribution to avoid ineffective scanning of clean areas.

[0139] For example, the laser scanning path in this embodiment adopts a spiral scanning path, which satisfies the following expression:

[0140] in, The scan radius is... The starting radius, The termination radius, For the number of scan cycles, The polar angle is denoted as . The values ​​for this embodiment are as follows: , , angular velocity Through this scanning path, the laser beam can cover the surface of the temperature measuring optical window, and the scanning speed v=ω·r is low in the center and high at the edges, ensuring more thorough cleaning of the central area.

[0141] During the cleaning process, auxiliary optical path data is acquired in real time, and the real-time transmittance data of the current area is determined based on this data. When the real-time transmittance data recovers to the target transmittance, subsequent cleaning of the current area is stopped or skipped. Furthermore, the cleaning process can also incorporate feedback judgment based on the deposition index. For example, when the absolute value of the average deposition index within a preset time period is less than the cleanliness threshold of 0.02, laser cleaning is stopped.

[0142] By employing the above-mentioned scheme, this application determines the deposition degree based on the fusion deposition thickness and deposition type, enabling the contamination treatment strategy to simultaneously adapt to both the deposition layer thickness and the deposition material type, avoiding inaccurate judgments based solely on a single thickness threshold. Furthermore, this application implements a temperature compensation strategy when the deposition degree reaches the temperature compensation threshold but not the cleaning threshold, correcting temperature measurement deviations caused by window deposition without cleaning, maintaining continuous and reliable temperature measurement results; and implements an in-situ cleaning strategy when the deposition degree reaches the cleaning threshold and the process is currently in a cleanable stage, reducing the interference of cleaning actions on the epitaxial growth stage and improving process continuity. Furthermore, this application employs different temperature compensation models for absorptive deposition and transparent thin film deposition, respectively compensating for temperature measurement errors caused by absorption attenuation and thin film reflection / interference, improving temperature compensation accuracy. Furthermore, this application determines the laser scanning path, dwell time, and scanning speed based on the contamination distribution map, enabling enhanced cleaning of severely contaminated areas and reduced laser action time in lightly contaminated areas, improving in-situ cleaning efficiency and reducing the risk of damage to the window substrate. Meanwhile, during the cleaning process, the transmittance is fed back in real time based on the auxiliary optical path data, and subsequent cleaning is stopped or skipped when the transmittance recovers to the target transmittance, which can realize closed-loop control of cleaning and avoid over-cleaning.

[0143] Optionally, if the process operation data also includes reaction chamber pressure and pressure change rate, the above-mentioned method for treating contamination of the temperature measuring optical window further includes: During the operation of semiconductor process equipment, a protective gas curtain is provided to the temperature measurement optical window according to the target purge flow rate. The target purge flow rate is obtained by dynamically adjusting the reference purge flow rate based on the reaction chamber pressure, pressure change rate and process stage information, so as to match the protective gas curtain with the current process state.

[0144] Specifically, during the operation of semiconductor process equipment, this application continuously provides a protective gas curtain to the temperature-sensing optical window to suppress the diffusion of reaction precursors, byproducts or particulate matter onto the surface of the temperature-sensing optical window.

[0145] For example, this application employs a digital PID algorithm to continuously acquire the reaction chamber pressure according to a sampling period of 200ms. The rate of pressure change is calculated based on the reaction chamber pressure at adjacent time points. Then, the reference purging flow rate is adjusted according to the pressure change rate to obtain the feedforward purging flow rate. Its expression is:

[0146] in, Based on the purging flow rate, This is the feedforward compensation coefficient.

[0147] Furthermore, based on the set value of the reaction chamber pressure and reaction chamber pressure Determine pressure deviation Its expression is:

[0148] Based on pressure deviation Perform PID feedback regulation to obtain the feedback regulation value. Its expression is:

[0149] in, , , These are the proportional coefficient, integral coefficient, and differential coefficient, respectively. ; Indicates pressure deviation The integral term accumulated over time.

[0150] Finally, based on the feedforward purging flow rate and feedback adjustment amount Obtain the target purge flow rate Its expression is:

[0151] Among them, process stage information is used to correct the target purge flow rate. For example, during the pressurization, depressurization, material change or cooling stages, the response weight to the rate of pressure change is increased; during the stable growth stage, the protective gas curtain is kept stable to reduce the disturbance to the process environment caused by drastic changes in purge flow rate.

[0152] Optionally, the above-mentioned method for treating contamination of the temperature-measuring optical window further includes: after completing in-situ cleaning of the temperature-measuring optical window, recording the updated transmittance benchmark value of the temperature-measuring optical window in a clean state; determining whether the temperature-measuring optical window has experienced substrate aging or irreversible damage based on the changing trend of multiple updated transmittance benchmark values; and, if it is determined that the temperature-measuring optical window has experienced substrate aging or irreversible damage, adjusting the temperature compensation threshold, cleaning threshold, and / or the benchmark parameters used to calculate transmittance data, and outputting a maintenance warning.

[0153] Specifically, the updated transmittance benchmark value is used to characterize the actual clean transmittance level that the temperature-sensing optical window can achieve after cleaning. If the temperature-sensing optical window is in a normal clean state after cleaning, the updated transmittance benchmark value is close to the initial clean transmittance benchmark value; if the temperature-sensing optical window suffers substrate aging, surface damage, or irreversible contamination, the updated transmittance benchmark value will continue to be lower than the historical benchmark level even after cleaning.

[0154] Furthermore, after each in-situ cleaning is completed, this application re-acquires auxiliary optical path data and calculates and updates the transmittance benchmark value. Subsequently, the updated transmittance baseline values ​​obtained after multiple cleaning processes were used to form a baseline value sequence: The system determines whether the temperature-measuring optical window has experienced substrate aging or irreversible damage based on the changing trend of the benchmark value sequence. For example, if the transmittance benchmark value shows a downward trend after multiple consecutive updates, or if the decrease in the transmittance benchmark value exceeds a preset decrease threshold after multiple consecutive updates, it is determined that the temperature-measuring optical window has experienced substrate aging or irreversible damage. The system then adjusts the temperature compensation threshold, cleaning threshold, and / or the benchmark parameters used to calculate the transmittance data, and outputs a maintenance warning.

[0155] By adopting the above scheme, this application can adapt to the actual aging state of the temperature measuring optical window and avoid inaccurate judgment of the degree of contamination caused by continuing to use the initial cleanliness standard.

[0156] Optionally, this application also performs predictive maintenance based on historical sedimentary indices. For example, a sedimentary growth trend is fitted based on historical sedimentary index data within the most recent cleaning intervals:

[0157] in, This represents the baseline value of the deposition index corresponding to the starting point of the fitting time. represents the average deposition growth rate; t represents the cumulative running time from the starting point of the fitting time.

[0158] Based on the current sedimentation index Cleaning threshold and average sediment growth rate The expression for predicting the next cleaning time is:

[0159] in, This indicates the current time when the prediction calculation is performed.

[0160] By adopting the above solution, this application can output maintenance suggestions before the temperature measuring optical window reaches the cleaning state, providing a basis for production scheduling and window maintenance.

[0161] The above-mentioned method for treating contamination in the temperature-measuring optical window calculates the deposition index based on the temperature-measuring main optical path data and determines the estimation result of the main optical path based on the deposition index. This method can reuse the original temperature measurement signals of the semiconductor process equipment to achieve online preliminary identification of the window deposition state, reduce reliance on additional monitoring hardware, and improve the real-time performance of deposition state monitoring.

[0162] Furthermore, this application calculates transmittance data based on auxiliary optical path data and uses a reference signal to normalize and compensate the transmission signal, which can reduce the impact of auxiliary monitoring light source power fluctuations, light source aging, or temperature drift on transmittance judgment and improve the reliability of window transmittance performance test results.

[0163] Furthermore, this application initiates multi-parameter Monte Carlo inversion only when the deposition index or transmittance data meet the preset inversion conditions. This enables the acquisition of deposition thickness and deposition optical parameters in scenarios where conventional models struggle to accurately determine conditions such as transparent films, absorptive deposition, and complex interference deposition, thereby improving the accuracy of identifying complex deposition states.

[0164] Furthermore, by performing multi-source integrated processing on the main optical path estimation results, auxiliary optical path estimation results, and inversion results, this application can determine the deposition thickness, deposition optical parameters, and deposition type based on the reliability of different data sources, thereby reducing the risk of temperature compensation errors or cleaning strategy errors caused by misjudgment of a single data source.

[0165] Furthermore, this application determines the contamination treatment strategy based on deposition state information and process stage information, which can compensate for the original temperature measurement value when the window deposition has not yet reached the cleaning conditions, and perform in-situ cleaning when the window deposition reaches the cleaning conditions and is in a cleanable process stage, thereby taking into account both temperature measurement accuracy and process continuity.

[0166] Furthermore, this application dynamically adjusts the purging flow rate of the protective air curtain during equipment operation, enabling the purging protection to match the reaction chamber pressure, pressure change rate, and process stage, thereby reducing the probability of precursors, byproducts, or particulate matter depositing on the surface of the temperature measuring optical window.

[0167] Furthermore, this application records updated transmittance benchmark values ​​after cleaning and determines the aging or irreversible damage of the window substrate based on the changing trends of multiple updated transmittance benchmark values. This enables long-term tracking of the temperature-measuring optical window's condition and avoids inaccurate judgment of the degree of contamination due to continued use of the initial benchmark.

[0168] In summary, this application enables online sensing of the contamination status of the temperature measurement optical window, accurate assessment of complex deposition states, real-time compensation for temperature measurement deviations, process-coordinated control of cleaning timing, and long-term maintenance judgment of the window status. This improves the accuracy and stability of temperature measurement results in semiconductor epitaxial processes, reduces the impact of window contamination on process control, and enhances the continuous operation capability of semiconductor process equipment. It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0169] Based on the same inventive concept, this application also provides a contamination treatment system for a temperature-measuring optical window. The system is installed in a semiconductor process equipment and is applicable to the above-mentioned contamination treatment method for the temperature-measuring optical window. The solution provided by the system is similar to the solution described in the above-mentioned method. Therefore, the specific limitations of one or more system embodiments provided below can be found in the limitations of the method above, and will not be repeated here.

[0170] Please see Figure 3 In one embodiment, the contamination treatment system for the temperature measurement optical window includes: an acquisition module, a calculation module, an inversion module, a comprehensive processing module, and a strategy determination module.

[0171] The acquisition module is used to acquire temperature measurement main optical path data, auxiliary optical path data, and process operation data; the process operation data includes process stage information. The calculation module is used to calculate the deposition index based on the temperature measurement main optical path data, and determine the main optical path estimation result based on the deposition index; the main optical path estimation result includes the thickness estimation result and the type prediction result; and the transmittance data is calculated based on the auxiliary optical path data. The inversion module is used to perform multi-parameter Monte Carlo inversion based on the temperature measurement main optical path data and auxiliary optical path data, provided that at least one of the deposition index and transmittance data meets the preset inversion conditions, and to obtain the inversion results; the inversion results include the inverted deposition thickness and the inverted deposition optical parameters. The integrated processing module is used to perform multi-source integrated processing based on the main optical path estimation results, transmittance data and inversion results to determine the deposition state information of the thermometric optical window; wherein, the deposition state information includes at least one of the following: fused deposition thickness, fused deposition optical parameters and deposition type; The strategy determination module is used to determine the contamination treatment strategy based on the deposition state information and process stage information; the contamination treatment strategy includes at least one of the temperature compensation strategy and the in-situ cleaning strategy.

[0172] The aforementioned contamination treatment system for the temperature measurement optical window can realize online sensing of the contamination status of the temperature measurement optical window, accurate assessment of complex deposition status, real-time compensation of temperature measurement deviation, process-coordinated control of cleaning timing, and long-term maintenance judgment of window status. This improves the accuracy and stability of temperature measurement results in semiconductor epitaxial processes, reduces the impact of window contamination on process control, and enhances the continuous operation capability of semiconductor process equipment. Each module in the contamination treatment system of the aforementioned temperature-measuring optical window can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0173] In one feasible embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 4 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements the aforementioned method for treating contamination of the temperature-measuring optical window. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0174] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0175] In one feasible embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described method for treating contamination of the temperature-measuring optical window.

[0176] In one feasible embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the above-described method for treating contamination of the temperature-measuring optical window.

[0177] In one feasible embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described method for treating contamination of the temperature-measuring optical window.

[0178] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0179] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for treating contamination in a temperature-measuring optical window, characterized in that, The method, applicable to semiconductor process equipment including a temperature-sensing optical window, comprises: Acquire temperature measurement main optical path data, auxiliary optical path data, and process operation data; The process operation data includes process stage information; Based on the temperature measurement main optical path data, the deposition index is calculated, and based on the deposition index, the main optical path estimation result is determined; The main optical path estimation results include thickness estimation results and type prediction results; Based on the auxiliary optical path data, the transmittance data is calculated; If at least one of the deposition index and the transmittance data satisfies a preset inversion condition, a multi-parameter Monte Carlo inversion is performed based on the temperature measurement main optical path data and the auxiliary optical path data to obtain the inversion result; the inversion result includes the inverted deposition thickness and the inverted deposition optical parameters. Based on the main optical path estimation results, the transmittance data, and the inversion results, multi-source integrated processing is performed to determine the deposition state information of the thermometric optical window; wherein, the deposition state information includes at least one of fused deposition thickness, fused deposition optical parameters, and deposition type; Based on the deposition state information and the process stage information, a contamination treatment strategy is determined; the contamination treatment strategy includes at least one of a temperature compensation strategy and an in-situ cleaning strategy.

2. The method according to claim 1, characterized in that, The temperature measurement main optical path data includes a first radiation intensity signal corresponding to the first temperature measurement wavelength and a second radiation intensity signal corresponding to the second temperature measurement wavelength. The deposition index is calculated based on the temperature measurement main optical path data, including: The original temperature value is calculated based on the first radiation intensity signal and the second radiation intensity signal; Calculate the original intensity ratio at the current moment based on the first radiation intensity signal and the second radiation intensity signal; Based on the original temperature measurement value, the corresponding cleanliness intensity ratio is determined from the pre-established cleanliness baseline relationship; the cleanliness baseline relationship is used to characterize the correspondence between different temperature measurement values ​​and the corresponding cleanliness intensity ratio under clean temperature measurement optical window conditions; Based on the original strength ratio and the clean strength ratio, determine the relative deviation information; The relative deviation information is filtered to obtain the deposition index.

3. The method according to claim 2, characterized in that, The step of determining the main optical path estimation result based on the deposition index includes: The variation characteristics of the sedimentation index are determined based on the sedimentation index at different times; Based on the deposition index and the variation characteristics of the deposition index, a type prediction result is determined; wherein, the type prediction result includes at least one of absorptive deposition, transparent thin film deposition, and complex deposition; If the predicted type is absorptive deposition, the thickness estimation result is determined based on the deposition index and the pre-established empirical model of deposition thickness. The type prediction result and / or the thickness estimation result are determined as the main optical path estimation result.

4. The method according to claim 1, characterized in that, The auxiliary optical path data includes the transmission signal acquired after the auxiliary monitoring beam passes through the temperature measuring optical window, and the reference signal used to characterize the output intensity of the auxiliary monitoring light source; The step of calculating transmittance data based on the auxiliary optical path data includes: The transmitted signal and the reference signal are preprocessed to obtain the transmitted signal value and the reference signal value; Acquire the reference transmission signal value and reference reference signal value under clean conditions; The transmittance data is calculated based on the transmitted signal value, the reference signal value, the baseline transmitted signal value, and the baseline reference signal value, and its expression is: in, For transmittance data, This is the transmitted signal value. As a reference signal value, As the reference transmission signal value, This is the reference signal value.

5. The method according to claim 1, characterized in that, The methods for determining whether the preset inversion conditions are met include: If at least one of the following conditions is met: the magnitude of change of the deposition index is greater than a preset magnitude threshold, the deposition index oscillates a preset number of times within a preset time period, the model uncertainty corresponding to the deposition index is greater than a preset uncertainty threshold, or the deviation between the transmittance data and the transmittance predicted based on the temperature measurement main optical path data is greater than a preset deviation threshold, then the preset inversion conditions are considered to be met.

6. The method according to claim 5, characterized in that, When at least one of the deposition index and the transmittance data satisfies a preset inversion condition, a multi-parameter Monte Carlo inversion is performed based on the temperature measurement main optical path data and the auxiliary optical path data to obtain the inversion result, including: Under the condition of satisfying the preset inversion, the first transmittance data at at least one temperature measurement wavelength is determined according to the temperature measurement main optical path data, and the second transmittance data at the auxiliary wavelength is determined according to the auxiliary optical path data. After combination, the measured transmittance data at multiple wavelengths are obtained. Using deposition thickness and deposition optical parameters as inversion parameters, a multi-parameter inversion model is constructed to characterize the deposition layer on the surface of the thermometric optical window; wherein, the deposition optical parameters include refractive index and extinction coefficient; Monte Carlo photon transmission simulation was performed based on the multi-parameter inversion model to obtain theoretical transmittance data corresponding to different inversion parameters. Based on the degree of matching between the measured transmittance data and the theoretical transmittance data, a likelihood function is constructed, and Bayesian sampling is performed on the parameters to be inverted based on the likelihood function to obtain the inversion result.

7. The method according to claim 6, characterized in that, The step of performing Bayesian sampling on the parameters to be inverted based on the likelihood function to obtain the inversion result includes: A prior distribution is set for the parameters to be inverted; a candidate parameter group is generated based on the prior distribution; the acceptance probability of the candidate parameter group is determined based on the difference between the theoretical transmittance data and the measured transmittance data corresponding to the candidate parameter group; the candidate parameter group is iteratively sampled according to the acceptance probability; and the inversion result is determined based on the sampling results after the convergence condition is met.

8. The method according to claim 1, characterized in that, The step of performing multi-source synthesis processing based on the main optical path estimation result, the transmittance data, and the inversion result to determine the deposition state information of the thermometric optical window includes: Based on the transmittance data, the auxiliary optical path estimation result is determined; the auxiliary optical path estimation result includes the auxiliary deposition thickness and the auxiliary deposition optical parameters; Based on the main optical path estimation result, the auxiliary optical path estimation result, and the inversion result, a thickness fusion process is performed to obtain the fused deposition thickness; Based on the transmittance data and the inversion results, the optical parameters for fusion deposition are determined; The type prediction result is corrected based on the variation characteristics of the fused deposition optical parameters and the deposition index to obtain the deposition type; the variation characteristics are determined based on the deposition index at different times.

9. The method according to claim 8, characterized in that, The step of performing thickness fusion processing based on the main optical path estimation result, the auxiliary optical path estimation result, and the inversion result to obtain the fused deposition thickness includes: Based on the main optical path estimation result, the auxiliary optical path estimation result, and the reliability of the inversion result, their respective fusion weights are determined; wherein, if the inversion result does not exist or the reliability of the inversion result does not meet the preset reliability condition, the fusion weight corresponding to the inversion result is set to zero; The thickness fusion process is performed using the fusion weight to obtain the fused deposition thickness.

10. The method according to claim 8, characterized in that, The step of determining the fusion deposition optical parameters based on the transmittance data and the inversion results includes: Based on the transmittance data and the preset thin film optical model, parameter inversion is performed to determine auxiliary optical parameters that match the theoretical transmittance with the transmittance data; If the inversion result is not available or the inversion result does not meet the preset confidence level, the auxiliary optical parameter will be determined as the fusion deposition optical parameter. If the inversion results meet the preset confidence level, the inversion optical parameters in the inversion results are determined as the fusion deposition optical parameters.

11. The method according to claim 1, characterized in that, The step of determining the pollution treatment strategy based on the deposition state information and the process stage information includes: The deposition degree of the thermometric optical window is determined based on the fusion deposition thickness and the deposition type. If the deposition level reaches the temperature compensation threshold but does not reach the cleaning threshold, it is determined that the contamination treatment strategy includes a temperature compensation strategy. When the deposition level reaches the cleaning threshold and the process stage information indicates that the process is currently in a cleanable stage, the contamination treatment strategy is determined to include an in-situ cleaning strategy. If the deposition level reaches the cleaning threshold and the process stage information indicates that the process is not currently in a cleanable stage, the contamination treatment strategy is determined to include a temperature compensation strategy.

12. The method according to claim 11, characterized in that, The temperature compensation strategy includes: The temperature compensation amount is determined based on the mapping model between the deposition state information and the temperature deviation; The original temperature value output by the main optical path of temperature measurement is corrected according to the temperature compensation amount to obtain the corrected temperature value. The mapping model is a nonlinear model established based on calibration data. The nonlinear model is used to characterize the temperature measurement deviation corresponding to different deposition state information. The nonlinear model includes at least one of a polynomial model, a lookup table model, or a piecewise fitting model.

13. The method according to claim 11, characterized in that, The in-situ cleaning strategy includes: The temperature-measuring optical window is scanned at multiple points to generate a contamination distribution map; The laser scanning path and the laser dwell time and / or scanning speed corresponding to each scanning area are determined according to the pollution distribution map. The area with a pollution level greater than a preset threshold is denoted as the first area, and the area with a pollution level less than or equal to the preset threshold is denoted as the second area. The laser dwell time corresponding to the first area is greater than the laser dwell time corresponding to the second area, and / or the scanning speed corresponding to the first area is less than the scanning speed corresponding to the second area. During the cleaning process, the auxiliary optical path data is acquired in real time, and the real-time transmittance data of the current area is determined based on the auxiliary optical path data. When the real-time transmittance data recovers to the target transmittance, the subsequent cleaning of the current area is stopped or skipped.

14. The method according to claim 1, characterized in that, The process operation data also includes reaction chamber pressure and pressure change rate; the method further includes: During the operation of the semiconductor process equipment, a protective gas curtain is provided to the temperature-measuring optical window according to a target purge flow rate; The target purging flow rate is obtained by dynamically adjusting the baseline purging flow rate based on the reaction chamber pressure, the pressure change rate, and the process stage information, so as to match the protective gas curtain with the current process state.

15. The method according to claim 1, characterized in that, The method further includes: After the temperature measuring optical window is cleaned in situ, the updated transmittance benchmark value of the temperature measuring optical window in the clean state is recorded. Based on the changing trends of multiple updated transmittance benchmark values, it is determined whether the temperature measuring optical window has experienced substrate aging or irreversible damage. If it is determined that the temperature-measuring optical window has experienced substrate aging or irreversible damage, the temperature compensation threshold, cleaning threshold, and / or the reference parameters used to calculate the transmittance data are adjusted, and a maintenance warning is output.

16. A contamination treatment system for a temperature-measuring optical window, characterized in that, The system, located within a semiconductor process apparatus including a temperature-measuring optical window, comprises: The acquisition module is used to acquire temperature measurement main optical path data, auxiliary optical path data, and process operation data; the process operation data includes process stage information. The calculation module is used to calculate the deposition index based on the temperature measurement main optical path data, and determine the main optical path estimation result based on the deposition index; the main optical path estimation result includes the thickness estimation result and the type prediction result; and to calculate the transmittance data based on the auxiliary optical path data. The inversion module is used to perform multi-parameter Monte Carlo inversion based on the temperature measurement main optical path data and the auxiliary optical path data, under the condition that at least one of the deposition index and the transmittance data meets a preset inversion condition, to obtain the inversion result; the inversion result includes the inverted deposition thickness and the inverted deposition optical parameters; The integrated processing module is used to perform multi-source integrated processing based on the main optical path estimation result, the transmittance data, and the inversion result to determine the deposition state information of the thermometric optical window; wherein, the deposition state information includes at least one of fused deposition thickness, fused deposition optical parameters, and deposition type; The strategy determination module is used to determine a contamination treatment strategy based on the deposition state information and the process stage information; the contamination treatment strategy includes at least one of a temperature compensation strategy and an in-situ cleaning strategy.

17. A semiconductor process apparatus, characterized in that, The device includes: A reaction chamber is used to provide an environment for semiconductor epitaxial processes; An optical window assembly includes a temperature-measuring optical window, which is disposed at the temperature-measuring channel of the reaction chamber and is used to provide a light signal transmission channel. The temperature measurement main optical path assembly is used to receive the radiation signal in the reaction cavity transmitted through the temperature measurement optical window and generate temperature measurement main optical path data. An auxiliary optical path assembly is used to emit an auxiliary monitoring beam and collect auxiliary optical path data obtained after the auxiliary monitoring beam passes through the temperature measuring optical window; A purge assembly, disposed on one side of the temperature measuring optical window, is used to provide a protective air curtain to the temperature measuring optical window; A cleaning module is used to perform in-situ cleaning of the temperature-measuring optical window based on a contamination treatment strategy. The contamination treatment system for the temperature measurement optical window as described in claim 16 is used to determine a contamination treatment strategy based on the temperature measurement main optical path data, auxiliary optical path data, and process operation data.

18. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1-15.

19. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1-15.

20. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-15.

Images