Film thickness monitoring method based on optical signal modulation and fiber coupling

Abstract

The application discloses a film thickness monitoring method based on optical signal modulation and optical fiber coupling and relates to the technical field of optical film monitoring. The method periodically modulates the detection light parameters and light transmission state in the film plating process, collects the return light signals after the action of the to-be-measured film layer, and respectively distributes the return light signals to a first collection channel and a second collection channel. Based on the modulation period corresponding to the detection light parameters, the monitoring wavelength light signal and the full spectrum light signal are synchronously segmented. Based on the synchronous sampling data, the spectral response change characteristics of the to-be-measured film layer at different wavelengths are extracted, the refractive index of the to-be-measured film layer is fitted, and then the corresponding film thickness monitoring algorithm update parameters are generated according to the real-time refractive index. The film thickness calculation mapping relationship in the preset film thickness calculation algorithm is updated online, the real-time optical film thickness of the to-be-measured film layer is calculated, and the problems of the increased film thickness calculation error and the increased plating stop judgment error in the prior art are solved.

Description

Technical Field

[0001] This invention relates to the field of optical film monitoring technology, and in particular to a film thickness monitoring method based on optical signal modulation and optical fiber coupling. Background Technology

[0002] In current thin film fabrication processes, film thickness monitoring is typically achieved through optical detection. This involves illuminating the film under test with detection light and collecting reflected, transmitted, or interferometric signals. The film thickness is then calculated based on changes in intensity, phase, spectral response, or interference period of the optical signals. In processes such as vacuum deposition, the film is usually located inside a vacuum chamber, while the light source, photodetector, and signal processing unit are often located outside the chamber or in a relatively independent detection area. The detection light needs to be transmitted through a viewing window, optical fiber, collimated optical path, coupling optical path, and return light acquisition optical path. Due to factors such as chamber sealing, spatial distance, temperature changes, umbrella or substrate movement, and stray light interference, the returned light signal is prone to problems such as amplitude attenuation, beam deflection, background light superposition, or noise enhancement. To improve the distinction between the detection signal and background light, stray light, and electrical noise signals, existing film thickness monitoring systems typically perform intensity modulation, pulse modulation, wavelength modulation, or phase modulation on the detection light and perform photoelectric conversion, amplification, filtering, and demodulation processing on the returned light signal to obtain monitoring data corresponding to changes in film thickness.

[0003] For example, in a vacuum coating scenario, a film thickness monitoring system typically includes a modulation light source, optical fiber, fiber optic coupler, detection optical path, photodetector, and film thickness calculation unit. The modulation light source generates detection light with a preset intensity, wavelength, pulse, or phase variation pattern. This detection light is transmitted via optical fiber and fiber optic coupler to the surface of the film under test or the monitoring substrate. During deposition, the film under test reflects, transmits, or interferes with the detection light, generating a return light signal. This return light signal is transmitted via the return light acquisition optical path and optical fiber to the photodetector, and after signal processing, forms optical monitoring data for film thickness calculation. The film thickness calculation unit calculates the real-time film thickness or determines the corresponding stop-deposition position based on the intensity, phase, spectral, or interference period changes of the return light signal.

[0004] For example, Chinese invention patent application CN116907363A discloses a multi-monitoring point optical film thickness direct monitoring system, including: a multi-point monitoring sheet, a light source device, an optical path transmission system, a spectral analysis device, a detector, an amplifier device, a signal conversion device, and a signal processing system. The light emitted by the light source device passes through the multi-point monitoring sheet and is then transmitted to the spectral analysis device for optical signal processing via the optical path transmission system. The detector converts the processed optical signal into a current signal and transmits it to the amplifier device. The amplifier device converts the current signal into a voltage signal and amplifies it. The amplified voltage signal is then converted into a digital signal and transmitted to the signal processing system. After signal processing, film thickness monitoring data is obtained.

[0005] However, in existing optical film thickness monitoring processes, film thickness calculation typically requires obtaining refractive index parameters and combining them with changes in the reflection, transmission, or interference of the detection light to calculate the film thickness. For optical thin films, the detection light signal is not only related to the physical thickness of the film but also to the film's refractive index, absorption characteristics, and changes in optical constants during the deposition process. In actual deposition, factors such as the deposition rate, substrate temperature, and material evaporation state can all cause changes in the film's refractive index, leading to deviations between the actual deposited film's refractive index and the preset refractive index. Especially when monitoring with a single wavelength or a limited number of wavelengths, the optical response within a finite band is insufficient to fully characterize the film's refractive index changes at different wavelengths. When the refractive index parameter is not corrected in time during the deposition process, dynamic changes in the refractive index are difficult to characterize in a timely manner, and the refractive index parameter does not match the real-time film thickness calculation model, resulting in increased film thickness calculation errors and errors in determining when to stop deposition. Summary of the Invention

[0006] To address the aforementioned technical problems in existing technologies, embodiments of the present invention provide a film thickness monitoring method based on optical signal modulation and optical fiber coupling. The technical solution is as follows:

[0007] A film thickness monitoring method based on optical signal modulation and fiber coupling is provided, which includes:

[0008] S1 periodically modulates the detection optical parameters and light transmission state during the thin film deposition process, transmits them to the film layer under test after optical signal coupling processing, collects the return optical signal after the film layer under test is applied, and distributes them to the first acquisition channel and the second acquisition channel respectively. The first acquisition channel is used to collect the monitoring wavelength optical signal, and the second acquisition channel is used to collect the full spectrum optical signal.

[0009] S2, based on the modulation period corresponding to the detection optical parameters, synchronously segment the monitoring wavelength optical signal and the full spectrum optical signal to obtain the monitoring wavelength sampling data and the full spectrum sampling data within the same modulation period, so as to establish the synchronous sampling data corresponding to the monitoring of the thickness of the film layer to be measured.

[0010] S3. Based on synchronous sampling data, extract the spectral response variation characteristics of the film under test at different wavelengths, and fit the refractive index of the film under test to obtain the real-time refractive index of the film under test.

[0011] S4. Generate corresponding film thickness monitoring algorithm update parameters based on real-time refractive index, and update the film thickness calculation mapping relationship in the preset film thickness calculation algorithm online to obtain the updated film thickness calculation algorithm.

[0012] S5 inputs the real-time refractive index and synchronous sampling data into the updated film thickness calculation algorithm to calculate the real-time optical film thickness of the film to be measured, and generates film thickness monitoring results based on the comparison results between the real-time optical film thickness and the target optical film thickness. When the film thickness monitoring results meet the preset stop-deposition conditions, a stop-deposition signal is output.

[0013] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0014] 1. By coordinating the periodic modulation of optical parameters and light transmission status, coupled optical signal transmission, synchronous acquisition through dual acquisition channels, full-spectrum refractive index fitting, and online updates of the film thickness calculation algorithm, the film thickness monitoring process no longer relies solely on changes in light intensity at a single monitoring wavelength. Instead, it combines the monitoring wavelength optical signal and the full-spectrum optical signal under the same modulation period reference for joint processing. Specifically, the first acquisition channel obtains monitoring wavelength sampling data directly related to film thickness determination, while the second acquisition channel obtains full-spectrum sampling data reflecting changes in response at different wavelengths. These two channels are synchronously segmented to form synchronous sampling data, which is then used to extract spectral response change features and fit the real-time refractive index. Subsequently, the real-time refractive index is used to update the film thickness calculation mapping relationship in the preset film thickness calculation algorithm, allowing the correspondence between refractive index parameters, optical thickness conversion relationships, monitoring wavelength response, and film thickness to be adjusted during the deposition process. Compared to methods that use only a fixed refractive index or a single-channel monitoring signal for film thickness calculation, this method can better reflect the changes in the optical state of the film during the actual deposition process, reduce the cumulative effects of refractive index shift, changes in optical path response, and single-wavelength judgment, thereby improving the consistency between the film thickness calculation results and the actual film state, and enabling the stop-deposition judgment to have better real-time adaptability.

[0015] 2. By dividing a single modulation period into a film detection time period, a reference sampling time period, and a background sampling time period, the return light signal, reference light signal, and background light signal can be obtained separately within the same modulation period. A temporal correspondence between the sampling data is established using period identifiers. This ensures that the monitoring wavelength light signal and the full-spectrum light signal are no longer used as independent acquisition results, but are synchronously captured according to the same period number, the same time period boundary, and the same sampling sequence. After synchronous segmentation, stable sampling areas are extracted from the monitoring wavelength segmented data and the full-spectrum segmented data. Transitional sampling points generated during detection light state switching are removed, and the representative light intensity value is determined based on the sampling values ​​within the stable sampling area. Then, background subtraction and reference normalization are performed to obtain the monitoring wavelength sampling data and the initial full-spectrum sampling data. Furthermore, the channel response correction parameter is determined based on the correspondence between the monitoring wavelength sampling data and the initial full-spectrum sampling data at the monitoring wavelength, and this parameter is used to correct the initial full-spectrum sampling data. The response difference between the first and second acquisition channels can be corrected within the same modulation period, enabling better channel consistency between the full-spectrum sampling data and the monitoring wavelength sampling data, thereby providing a more stable data basis for subsequent refractive index fitting and film thickness calculation.

[0016] 3. By converting the real-time refractive index into update parameters for the film thickness monitoring algorithm, and using these update parameters to update the film thickness calculation mapping relationship in the preset film thickness calculation algorithm online, the film thickness calculation and coating stop judgment can be adjusted according to the current optical state of the film layer. Based on the difference between the real-time refractive index and the refractive index reference value, refractive index offset parameters, optical thickness conversion correction parameters, and coating stop judgment correction parameters are generated. Then, based on the real-time refractive index and the film thickness estimation range corresponding to the current modulation period, a candidate film thickness sequence is constructed, generating first mapping data between candidate thicknesses and candidate optical thicknesses, and second mapping data between candidate thicknesses and theoretical response data of the monitoring wavelength. The updated film thickness calculation algorithm can simultaneously include the conversion relationship between the refractive index parameter and optical thickness, the correspondence between the monitoring wavelength sampling data and the film thickness, and the correlation between optical thickness deviation and coating stop judgment conditions. In the real-time calculation stage, the monitoring wavelength sampling data corresponding to the current modulation period is matched with the second mapping data to determine the actual thickness of the film to be measured, and the real-time optical film thickness is calculated in conjunction with the real-time refractive index. By combining online updates with predictive judgments, film thickness monitoring results can better serve the determination of when to stop coating, reducing coating stop deviations caused by changes in refractive index, response mapping lag, or modulation cycle delay. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating the film thickness monitoring method based on optical signal modulation and optical fiber coupling provided in an embodiment of the present invention;

[0019] Figure 2 A flowchart corresponding to periodic modulation and dual-channel allocation is provided for embodiments of the present invention;

[0020] Figure 3 A flowchart illustrating the synchronization segmentation and channel response correction provided in this embodiment of the invention;

[0021] Figure 4 The flowchart below shows the real-time refractive index fitting and film thickness algorithm update provided in the embodiments of the present invention. Detailed Implementation

[0022] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0023] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0024] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0025] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0026] The existing direct monitoring process for optical film thickness typically involves: generating detection light from a light source, which is then collimated, focused, passed through a viewing window, a monitoring plate, or the film to be measured before entering an optical fiber. The wavelength is then selected by a monochromator or spectral analysis device, followed by photoelectric conversion by a detector. After amplification, filtering, and data acquisition, the data is input into a computer to calculate the changes in transmittance or light intensity. Algorithms such as photoelectrode method, previous extreme value method, and backtracking method are used to determine the film thickness or the point where coating stops. The main focus is on improving the light signal intensity, coupling efficiency, and signal-to-noise ratio.

[0027] This invention organizes the detected optical parameters and light transmission state according to the same modulation period, generating a return optical signal, a reference optical signal, and a background optical signal within each modulation period. Then, through optical signal coupling processing, the optical signal after the coating layer is applied is distributed to a first acquisition channel and a second acquisition channel. The first acquisition channel is used for acquiring the monitoring wavelength optical signal, and the second acquisition channel is used for acquiring the full-spectrum optical signal. Compared with existing single-channel or single-transmittance curve-based stop-deposition processes, this invention does not simply add an optical path or spectrometer, but binds the dual-channel data to the same modulation period and performs synchronous segmentation, background subtraction, reference normalization, and channel response correction, thereby forming synchronous sampling data that can be used for film thickness calculation and refractive index fitting. Subsequently, based on the full-spectrum sampling data, the spectral response variation characteristics at different wavelengths are extracted, the real-time refractive index of the coating layer is fitted, and the real-time refractive index is used to update the film thickness calculation mapping relationship and the stop-deposition judgment condition online. Therefore, the main difference between this invention and the prior art lies in the combined processing of dual-channel synchronous acquisition, same-period signal correction, full-spectrum refractive index fitting and online updating of film thickness calculation algorithm, which enables the film thickness monitoring process to better adapt to the impact of refractive index changes and channel response changes on the monitoring results during the coating process.

[0028] This invention provides a film thickness monitoring method based on optical signal modulation and optical fiber coupling, such as... Figure 1 The flowchart shown is for a film thickness monitoring method based on optical signal modulation and fiber coupling. The processing flow of this method may include the following steps:

[0029] S1 periodically modulates the detection optical parameters and light transmission state during the thin film deposition process, transmits them to the film layer under test after optical signal coupling processing, collects the return optical signal after the film layer under test is applied, and distributes them to the first acquisition channel and the second acquisition channel respectively. The first acquisition channel is used to collect the monitoring wavelength optical signal, and the second acquisition channel is used to collect the full spectrum optical signal.

[0030] S2, based on the modulation period corresponding to the detection optical parameters, synchronously segment the monitoring wavelength optical signal and the full spectrum optical signal to obtain the monitoring wavelength sampling data and the full spectrum sampling data within the same modulation period, so as to establish the synchronous sampling data corresponding to the monitoring of the thickness of the film layer to be measured.

[0031] S3. Based on synchronous sampling data, extract the spectral response variation characteristics of the film under test at different wavelengths, and fit the refractive index of the film under test to obtain the real-time refractive index of the film under test.

[0032] S4. Generate corresponding film thickness monitoring algorithm update parameters based on real-time refractive index, and update the film thickness calculation mapping relationship in the preset film thickness calculation algorithm online to obtain the updated film thickness calculation algorithm.

[0033] S5 inputs the real-time refractive index and synchronous sampling data into the updated film thickness calculation algorithm to calculate the real-time optical film thickness of the film to be measured, and generates film thickness monitoring results based on the comparison results between the real-time optical film thickness and the target optical film thickness. When the film thickness monitoring results meet the preset stop-deposition conditions, a stop-deposition signal is output.

[0034] like Figure 2 The diagram shows a flowchart illustrating the periodic modulation and dual-channel allocation provided in this embodiment of the invention. In step S1, the film thickness monitoring parameters corresponding to the film layer under test are first acquired. These parameters refer to the process parameters and acquisition parameters used to determine the modulation method and sampling rhythm of the detection light, and include at least the monitoring wavelength, sampling frequency, and film deposition rate. The monitoring wavelength is read from the coating process file, which typically records the monitoring wavelength, target optical film thickness, and material type for each film layer under test. The sampling frequency is read from the acquisition configuration file. The film deposition rate is obtained from the deposition rate setting value in the current coating process file. When both real-time deposition rate and process-set deposition rate exist simultaneously, the real-time deposition rate is prioritized, and the process-set deposition rate is used as an anomaly verification reference.

[0035] After acquiring the film thickness monitoring parameters, the modulation period of the detection optical parameters is determined based on the film deposition rate and the preset allowable film thickness monitoring range. The preset allowable film thickness monitoring range is used to limit the maximum allowable optical thickness change range of the film under test within a single modulation period. If the film deposition rate is the physical thickness deposition rate, the refractive index reference value and incident angle parameter of the film under test are first read, and the physical thickness deposition rate is converted into the optical thickness change rate. Then, the optical thickness change within a single modulation period is obtained by multiplying the optical thickness change rate by the modulation period duration. If the film deposition rate is the optical thickness deposition rate, the optical thickness change within a single modulation period is directly calculated using this optical thickness deposition rate.

[0036] After determining the modulation period, each modulation period is divided into a film detection period, a reference sampling period, and a background sampling period. The film detection period refers to the time interval during which the detection light acts on the film under test and collects the light signal returned after the film's action. The reference sampling period refers to the time interval during which the detection light enters the reference optical path without passing through the deposition area of ​​the film under test and collects the reference light signal used to characterize the current incident light intensity. The background sampling period refers to the time interval during which the detection light is in a state of no effective incident light and collects the corresponding signals of background light, dark current, and electrical noise. The length of each of these three time intervals is determined by the minimum number of sampling points and sampling frequency required for each. These time intervals are combined in a preset order to form a complete modulation period. The preset order is stored in the modulation configuration file; for example, the film detection period can be executed first, followed by the reference sampling period, and finally the background sampling period. Alternatively, other orders can be set based on the actual optical path switching response time, but consistency must be maintained throughout the same batch of deposition processes.

[0037] Within each modulation cycle, the detection light parameters and light transmission state are controlled according to the modulation configuration file. Detection light parameters include at least one of detection light intensity, detection light pulse state, detection light wavelength, or detection light gating state. The light transmission state refers to the state in which the detection light switches between the film detection optical path, the reference optical path, and the state of no effective incident light. Specifically, during the film detection time period, the detection light is transmitted to the film under test after optical signal coupling processing, causing the film under test to reflect, transmit, or interfere with the detection light, and the return light signal is obtained by the acquisition end; during the reference sampling time period, the detection light enters the reference optical path, which is an optical path that does not pass through the deposition area of ​​the film under test and is used to characterize the current incident light intensity, and the reference light signal is obtained by the acquisition end; during the background sampling time period, the control unit shuts off the detection light source, blocks the detection light, or switches the optical path to the state of no effective incident light, and the background light signal is obtained by the acquisition end.

[0038] To ensure that subsequent dual-channel acquired data can be processed according to the same time reference, a period identifier is generated for each modulation cycle. A period identifier is a data marker used to distinguish different modulation cycles and indicate the position of each time period. It includes at least a cycle number, the start and end sampling points of the film detection time period, the start and end sampling points of the reference sampling time period, the start and end sampling points of the background sampling time period, and the end sampling point of the background sampling time period. The cycle number is generated incrementally according to the execution order of the modulation cycles. The return optical signal, reference optical signal, and background optical signal acquired within the same modulation cycle are bound to this period identifier, enabling subsequent steps to determine the temporal correspondence of each sampled data within the same modulation cycle based on the period identifier.

[0039] In one optional implementation, after completing periodic modulation, the return optical signal, reference optical signal, and background optical signal formed within the same modulation period are used as the optical signals to be assigned. The optical signals to be assigned refer to the sets of optical signals that need to be input into the first acquisition channel and the second acquisition channel, respectively. The first acquisition channel is used to acquire monitoring wavelength optical signals, which are optical signals obtained at the monitoring wavelength corresponding to the film layer under test. The second acquisition channel is used to acquire full-spectrum optical signals, which are optical signals obtained at multiple sampling wavelengths within a preset full-spectrum acquisition range. The preset full-spectrum acquisition range is jointly determined by the working band of the full-spectrum acquisition and the spectral response range of the film layer material under test.

[0040] Before optical path allocation, the first channel allocation coefficient and the second channel allocation coefficient are determined based on the monitoring wavelength acquisition requirements corresponding to the first acquisition channel and the full-spectrum acquisition requirements corresponding to the second acquisition channel. The monitoring wavelength acquisition requirements include the minimum effective light intensity, detector response range, and signal-to-noise ratio requirements at the monitoring wavelength; the full-spectrum acquisition requirements include the minimum effective light intensity of each sampling wavelength within the preset full-spectrum acquisition range, the spectrometer response range, and the data integrity requirements for refractive index fitting. The aforementioned minimum effective light intensity, detector response range, and spectrometer response range are obtained through the calibration process. The required light intensity ratio for the two acquisition channels is determined according to the channel calibration table, and this light intensity ratio is converted into the first channel allocation coefficient and the second channel allocation coefficient. The first channel allocation coefficient represents the proportion of light signal entering the first acquisition channel, and the second channel allocation coefficient represents the proportion of light signal entering the second acquisition channel; together, they determine the channel allocation relationship of the light signals to be allocated.

[0041] In actual allocation, the optical signals to be allocated are distributed according to the allocation coefficients of the first and second channels. The optical path allocation components employ beam splitters, beam splitting fiber bundles, convergent beam splitting multimode fibers, or equivalent optical signal allocation structures. When beam splitting fiber bundles are used, allocation is achieved through the ratio of the number of fibers or the cross-sectional area of ​​light transmitted to the two acquisition channels; when beam splitters are used, allocation is achieved through the transmission-reflection ratio of the beam splitter. To ensure data consistency within the same modulation period, the return optical signal, reference optical signal, and background optical signal all use the same optical path allocation relationship, ensuring that the three types of optical signals received by the first and second acquisition channels have the same periodic reference.

[0042] The optical signal entering the first acquisition channel is processed by a selected monitoring wavelength to form a monitoring wavelength optical signal. Monitoring wavelength selection can be achieved using a narrow-linewidth monochromator, bandpass filter, tunable filter, or a spectrometer with a specified wavelength readout method. The optical signal entering the second acquisition channel is acquired within a preset full-spectrum acquisition range to form a full-spectrum optical signal. Full-spectrum acquisition can be achieved using a full-spectrum analyzer or spectrometer. Based on the temporal correspondence of each time period in the period identifier, the acquisition results of the first and second acquisition channels are marked with time periods. Acquisition results whose sampling time falls within the film detection time period are marked as return light acquisition data; those whose sampling time falls within the reference sampling time period are marked as reference light acquisition data; and those whose sampling time falls within the background sampling time period are marked as background light acquisition data. Finally, the acquisition results of the first and second acquisition channels with the same period identifier are correlated to obtain the monitoring wavelength optical signal and the full-spectrum optical signal under the same modulation period reference, providing a data basis for subsequent synchronous segmentation, background subtraction, reference normalization processing, and refractive index fitting.

[0043] like Figure 3 The diagram shows a flowchart illustrating the synchronous segmentation and channel response correction provided in this embodiment of the invention. In step S2, the monitoring wavelength optical signal output from the first acquisition channel and the full-spectrum optical signal output from the second acquisition channel are received first. The sampling time corresponding to each sampled data is generated by a unified sampling clock. When the acquisition result only contains the sampling sequence number, the sampling time is calculated based on the sampling sequence number and the sampling frequency, thereby enabling the data from the two acquisition channels to be processed under the same time reference.

[0044] After acquiring the sampling time, the period assignment of the monitored wavelength optical signal and the full-spectrum optical signal is determined based on the period identifier: the sampling time or sampling point number of each sampled data is compared with the start and end sampling points in the period identifier; when a sampled data falls within the sampling point range corresponding to a certain period number, the sampled data is assigned to that period number; when the sampled data further falls within the sampling point range of the film layer detection time period, the reference sampling time period, or the background sampling time period, the sampled data is marked as data of the corresponding time period. This allows the determination of the period number and time period corresponding to each sampled data.

[0045] After determining the cycle assignment, monitoring wavelength optical signals and full-spectrum optical signals belonging to the same cycle number are identified as optical signals to be segmented within the same modulation cycle. Optical signals to be segmented refer to data sets that have been confirmed to belong to the same modulation cycle but have not yet been segmented according to the film detection time period, reference sampling time period, and background sampling time period. Synchronous truncation of the optical signals to be segmented is performed according to the time period boundaries recorded in the cycle identifier. The time period boundaries refer to the start and end sampling points corresponding to the film detection time period, reference sampling time period, and background sampling time period, respectively. Synchronous truncation means trunculating the data from the first and second acquisition channels using the same time period boundaries under the same cycle number, ensuring that the segmented data obtained from the two acquisition channels have the same modulation cycle reference and time period attributes. The resulting monitoring wavelength segmented data includes monitoring wavelength return segmented data, monitoring wavelength reference segmented data, and monitoring wavelength background segmented data; the full-spectrum segmented data includes full-spectrum return segmented data, full-spectrum reference segmented data, and full-spectrum background segmented data.

[0046] Stable sampling region extraction is performed on the monitoring wavelength segmented data and full-spectrum segmented data. This refers to the sampling interval within the same time period after removing the influence of detection light state switching. Since the light intensity may undergo short-term transitions when the detection light switches from the film detection state to the reference sampling state, or from the reference sampling state to the background sampling state, the representative light intensity value is not directly calculated using all sampling points from each time period. Instead, the number of transition sampling points is determined based on the pre-calibrated optical path switching response time and sampling frequency. The optical path switching response time can be obtained by collecting the light intensity stabilization time during the commissioning phase when the light source switches, the light-shielding component operates, or the optical path switching component operates. Based on the optical path switching response time and sampling frequency, the number of sampling points to be removed at the start and end positions of each time period is calculated, and the remaining sampling points after removal are used as the sampling points within the stable sampling region.

[0047] Within the stable sampling area, representative light intensity values ​​for each time period are determined based on the sampled values. These representative light intensity values ​​characterize the stable light intensity level for a given time period within the same modulation period. For segmented data of the monitoring wavelength, the average value of the sampled values ​​within the stable sampling area is calculated to obtain the returned light intensity value, the reference light intensity value, and the background light intensity value of the monitoring wavelength. For segmented full-spectrum data, the average light intensity within the stable sampling area is calculated for each sampling wavelength to obtain the returned full-spectrum light intensity value, the reference full-spectrum light intensity value, and the background full-spectrum light intensity value at different wavelengths.

[0048] In one optional implementation, after obtaining the representative light intensity value, background subtraction and reference normalization are performed on the monitoring wavelength segmented data and the full-spectrum segmented data. Background subtraction refers to using the representative light intensity value corresponding to the background sampling time period to remove the influence of background light, dark current, and electrical noise on the film detection data and reference sampling data. For the monitoring wavelength segmented data, the monitoring wavelength return light intensity value is subtracted from the monitoring wavelength background light intensity value to obtain the effective return light intensity value of the monitoring wavelength, and the monitoring wavelength reference light intensity value is subtracted from the monitoring wavelength background light intensity value to obtain the effective reference light intensity value of the monitoring wavelength. Reference normalization refers to using the effective reference light intensity value within the same modulation period as the incident light intensity reference to proportionally convert the effective return light intensity value. The ratio between the effective return light intensity value of the monitoring wavelength and the effective reference light intensity value of the monitoring wavelength is calculated to obtain the monitoring wavelength sampling data within that modulation period.

[0049] For full-spectrum segmented data, background subtraction and reference normalization are performed sequentially for each sampling wavelength. For any sampling wavelength, the full-spectrum returned light intensity value at that sampling wavelength is subtracted from the full-spectrum background light intensity value at that sampling wavelength to obtain the full-spectrum effective returned light intensity value at that sampling wavelength; the full-spectrum reference light intensity value at that sampling wavelength is subtracted from the full-spectrum background light intensity value at that sampling wavelength to obtain the full-spectrum effective reference light intensity value at that sampling wavelength. Before calculating the initial full-spectrum sampling value, the validity of the full-spectrum effective reference light intensity value at that sampling wavelength is determined: the full-spectrum effective reference light intensity value is used as a reference for the incident light intensity at that sampling wavelength. When the full-spectrum effective reference light intensity value is less than or equal to the preset reference light intensity lower limit, the ratio calculation at that sampling wavelength is not performed, and that sampling wavelength is marked as an invalid sampling wavelength. The preset reference light intensity lower limit is obtained through the calibration process, specifically by collecting reference light intensity data for each sampling wavelength under stable light source output and film-free deposition conditions, and determining it in conjunction with the background noise fluctuation range. This lower limit is used to prevent abnormal amplification of the initial full-spectrum sampling value due to an excessively small denominator when the reference light intensity is close to zero.

[0050] The ratio between the effective returned light intensity value and the effective reference light intensity value of the full spectrum is then calculated to obtain the initial full spectrum sampling value at that sampling wavelength. The above processing is performed sequentially on each sampling wavelength within the full spectrum acquisition range to form the initial full spectrum sampling data within the same modulation period.

[0051] Before generating the initial full-spectrum sampling data, each sampling wavelength undergoes validity screening: if a sampling wavelength lacks any of the returned light intensity value, reference light intensity value, or background light intensity value, it is marked as a missing sampling wavelength; if the original light intensity value of a sampling wavelength reaches the saturation threshold, it is marked as a saturated sampling wavelength; if the ratio between the effective returned light intensity value and the background noise fluctuation value of a sampling wavelength is lower than the preset signal-to-noise ratio threshold, it is marked as a low signal-to-noise ratio sampling wavelength. The saturation threshold is determined by the range calibration results, the background noise fluctuation value is determined by the sampling fluctuation range of the stable sampling area within the background sampling time period, and the preset signal-to-noise ratio threshold is determined by the historical calibration data and the data stability requirements for refractive index fitting.

[0052] Sampling points marked as invalid, missing, saturated, or low signal-to-noise ratio wavelengths are not directly used for refractive index fitting. If valid sampling wavelengths exist on both sides of the wavelength, and these wavelengths are adjacent to the wavelength in the wavelength ordering, interpolation can be performed based on the initial full-spectrum sampling values ​​of the two valid wavelengths to obtain a replacement sampling value, and an interpolation marker is added to this replacement value. If there are insufficient adjacent valid sampling wavelengths for interpolation, or if multiple adjacent sampling wavelengths are marked as invalid, the wavelength region is removed from the refractive index fitting data of the current modulation period. If the number of valid sampling wavelengths after removal is lower than the preset lower limit for the number of valid wavelengths, the full-spectrum sampling data of the current modulation period is marked as not meeting the refractive index fitting conditions, and data is collected for the next modulation period. The preset lower limit for the number of valid wavelengths is determined by the full-spectrum acquisition range, the sampling wavelength interval, and the minimum wavelength coverage required for refractive index fitting, and is stored in the algorithm configuration file.

[0053] To reduce the impact of response differences between the two acquisition channels on subsequent refractive index fitting, channel response correction parameters are determined based on the correspondence between the monitoring wavelength sampling data and the initial full-spectrum sampling data at the monitoring wavelength. These parameters are used to correct for the response differences between the first and second acquisition channels at the same monitoring wavelength. First, a sampling wavelength identical to the monitoring wavelength is found in the initial full-spectrum sampling data, and its corresponding initial full-spectrum sampling value is read. If no sampling wavelength exactly matches the monitoring wavelength in the full-spectrum sampling data, the initial full-spectrum sampling values ​​corresponding to sampling wavelengths located on either side of and adjacent to the monitoring wavelength are read, and linear interpolation is performed based on the wavelength distance to obtain the initial full-spectrum sampling value at the monitoring wavelength. Then, the ratio between the monitoring wavelength sampling data and the initial full-spectrum sampling value at the monitoring wavelength is calculated, and this ratio is used as the channel response correction parameter for the same modulation period.

[0054] After obtaining the channel response correction parameter, the initial full-spectrum sampling data is corrected using this parameter. The initial full-spectrum sampling value corresponding to each sampling wavelength in the initial full-spectrum sampling data is multiplied by the channel response correction parameter to obtain the corrected full-spectrum sampling value. These corrected full-spectrum sampling values ​​are then combined according to the sampling wavelength order to obtain the full-spectrum sampling data within the modulation period. Therefore, the full-spectrum sampling data at the monitoring wavelength maintains a correspondence with the monitoring wavelength sampling data obtained by the first acquisition channel, reducing data offset caused by differences in detector response, optical path allocation ratio, or spectral acquisition sensitivity between the two acquisition channels.

[0055] By associating the monitoring wavelength sampling data, full-spectrum sampling data, and corresponding period identifiers within the same modulation period, synchronous sampling data corresponding to the monitoring of the film thickness under test is established. Synchronous sampling data refers to a data set formed by combining monitoring wavelength sampling data, full-spectrum sampling data, period numbers, time period boundaries, and sampling time references, using the same modulation period as the data unit. This synchronous sampling data includes both monitoring wavelength sampling data used for film thickness monitoring and full-spectrum sampling data used for subsequent refractive index fitting, and both have the same period identifier, thus providing a unified data foundation for subsequent extraction of spectral response change characteristics, fitting of real-time refractive index, and updating of the film thickness calculation algorithm.

[0056] During step S3, the full-spectrum acquisition range corresponding to the second acquisition channel is read. The full-spectrum sampling data output by the second acquisition channel typically includes multiple sampling wavelengths and the sampling value corresponding to each sampling wavelength. The full-spectrum sampling data within the current modulation period are arranged in ascending order of sampling wavelength to obtain the spectral response sequence corresponding to the current modulation period. The spectral response sequence refers to the set of spectral sampling values ​​arranged in wavelength order, used to represent the optical response distribution of the film under test within the full-spectrum acquisition range during the current modulation period.

[0057] After determining the spectral response sequence corresponding to the current modulation period, the initial spectral response sequence corresponding to the start of film deposition is obtained. The initial spectral response sequence is obtained during the stable sampling phase before the current film to be deposited. When the stable sampling phase before deposition is used as the initial reference, under the condition that the film to be deposited has not yet formed or the film thickness change is negligible, full-spectrum sampling data is generated according to step S2, and the full-spectrum sampling data is arranged in wavelength order to form the initial spectral response sequence.

[0058] like Figure 4The diagram shows a flowchart illustrating the real-time refractive index fitting and film thickness algorithm update provided in this embodiment of the invention. The spectral response sequence corresponding to the current modulation period is compared wavelength by wavelength with the initial spectral response sequence: at the same sampling wavelength, the difference between the spectral sample value of the current modulation period and the initial spectral sample value is calculated to obtain the spectral response change corresponding to that sampling wavelength. If a sampling wavelength in the current spectral response sequence is not completely consistent with the sampling wavelength in the initial spectral response sequence, interpolation is performed based on the sampling values ​​of adjacent sampling wavelengths to ensure that the two sets of data are compared at the same sampling wavelength. The spectral response change represents the degree of change in the spectral response of the film under test relative to its initial state in the current modulation period.

[0059] The trend of spectral response change is determined based on the difference in spectral response changes between adjacent wavelengths. Adjacent wavelengths refer to two sampling wavelengths that are adjacent in their sorted order in the spectral response sequence arranged from smallest to largest wavelength. For any two adjacent wavelengths, the difference in their corresponding spectral response changes is calculated, and the trend within that adjacent wavelength interval is determined based on the sign and magnitude of the difference. When the spectral response change corresponding to the later sampling wavelength is greater than that corresponding to the earlier sampling wavelength, the wavelength interval is marked as having an upward trend; when the spectral response change corresponding to the later sampling wavelength is less than that corresponding to the earlier sampling wavelength, the wavelength interval is marked as having a downward trend; when the difference is within a preset trend determination tolerance, the wavelength interval is marked as having a flat trend. The preset trend determination tolerance is determined within the range of repeated sampling fluctuations under stable light source conditions.

[0060] After the above processing is completed, based on the spectral response sequence, the spectral response changes at different wavelengths, the spectral response change trends, and the monitoring wavelength sampling data, a set of data describing the optical state of the film under test at different wavelengths is generated. This set includes the current spectral response sequence, the spectral response changes corresponding to each sampling wavelength, the spectral response change trends corresponding to adjacent wavelength intervals, and the monitoring wavelength sampling data within the same modulation period. This spectral response change feature is then used to fit the refractive index of the film under test.

[0061] In one optional implementation, a candidate range of refractive index is determined based on a reference value corresponding to the film layer under test. The reference value represents the refractive index of the film layer material under design conditions. The candidate range of refractive index refers to the range of refractive index values ​​allowed to be searched during real-time refractive index fitting. Preset refractive index lower and upper values ​​are read. The reference value is subtracted from the preset refractive index lower value to obtain the lower limit of the candidate range, and the reference value is added to the preset refractive index upper value to obtain the upper limit of the candidate range. The preset refractive index lower and upper values ​​can be determined based on historical plating data of similar materials, allowable deviations in material refractive index, or process calibration results.

[0062] After the candidate range of refractive indices is determined, a set of candidate refractive indices is generated within this range according to a preset refractive index interval. The set of candidate refractive indices refers to multiple refractive index values ​​used for fitting and comparison. The preset refractive index interval is determined by the required fitting accuracy and computational capability. Starting from the lower limit of the candidate refractive index range, the interval is gradually increased according to the preset refractive index interval until the upper limit of the candidate refractive index range is reached, thus obtaining the set of candidate refractive indices.

[0063] Based on the monitoring wavelength sampling data and the deposition process parameters of the film under test, the estimated range of film thickness corresponding to the current modulation cycle is determined. The deposition process parameters include at least the film deposition rate, the duration of the current modulation cycle, and the deposition time of the film under test. The film deposition rate can be monitored and output by crystal oscillation; the duration of the current modulation cycle is calculated by combining the start and end sampling points in the cycle identifier with the sampling frequency; the deposition time is calculated based on the start time of the film under test and the corresponding time in the current modulation cycle. A rough estimate of the current film thickness is calculated based on the deposition rate and the deposition time, and the estimated film thickness range is determined using this rough estimate as the center, combined with a preset thickness extension range. The preset thickness extension range is determined by the deposition rate fluctuation range and the allowable error.

[0064] Within the estimated film thickness range, the estimated film thickness corresponding to the current modulation period is determined based on a preset film thickness calculation algorithm. This preset algorithm is a transmittance calculation algorithm based on the thin film feature matrix. Several candidate film thicknesses are selected within the estimated range, and each candidate thickness is input into the preset algorithm to obtain the theoretical response data for the monitoring wavelength corresponding to each candidate thickness. Then, the theoretical response data for each monitoring wavelength is compared with the sampling data for the monitoring wavelength corresponding to the current modulation period, and the candidate film thickness with the smaller difference is determined as the estimated film thickness for the current modulation period. Therefore, the estimated film thickness is derived from the combined constraints of the monitoring wavelength sampling data and the deposition process parameters.

[0065] After obtaining the candidate refractive index set and the estimated film thickness, each candidate refractive index and the estimated film thickness are input into a preset film thickness calculation algorithm to calculate the theoretical spectral response data corresponding to each candidate refractive index. The theoretical spectral response data refers to the theoretical optical response at different sampling wavelengths calculated by the preset film thickness calculation algorithm under the given candidate refractive index and estimated film thickness. For each candidate refractive index in the candidate refractive index set, the candidate refractive index, the estimated film thickness, the substrate refractive index, the monitoring wavelength, and the sampling wavelengths within the full spectrum acquisition range are input into the preset film thickness calculation algorithm to calculate the theoretical response value corresponding to each sampling wavelength. These values ​​are then arranged in order of sampling wavelength to form the theoretical spectral response data corresponding to that candidate refractive index. The substrate refractive index is obtained from the film system design file or a material parameter library.

[0066] The theoretical spectral response data is compared with the spectral response sequence corresponding to the current modulation period to obtain the spectral response deviation for each candidate refractive index. The spectral response deviation refers to the difference between the theoretical response value and the measured response value at the same sampling wavelength. For each candidate refractive index, the difference between the theoretical response value and the corresponding sampled value in the current spectral response sequence is calculated at each sampling wavelength, and these differences are used as the spectral response deviation data for that candidate refractive index.

[0067] After determining the spectral response deviation, the deviation is judged to exceed limits based on the changes in spectral response at different wavelengths. The preset deviation threshold is obtained through the calibration process and can be determined jointly based on the noise from the stable light source, the fluctuation of repeated sampling by the spectrometer, and the allowable error of film thickness monitoring. For any candidate refractive index, the spectral response deviation corresponding to each sampling wavelength is compared with the preset deviation threshold. When the spectral response deviation at a certain sampling wavelength is greater than the preset deviation threshold, that sampling wavelength is marked as a sampling wavelength with excessive deviation.

[0068] After marking the wavelengths exceeding the deviation limit, the width of the over-limit band is determined based on these wavelengths. The over-limit band width refers to the wavelength span where the wavelengths exceeding the deviation limit are concentrated, given the sequential arrangement of the sampling wavelengths. Specifically, the wavelengths exceeding the deviation limit are checked in ascending order. When multiple wavelengths exceeding the deviation limit are adjacent in their sequence number, they are divided into the same over-limit band. For each over-limit band, the difference between the maximum and minimum sampling wavelengths within that band is calculated to obtain the width of the over-limit band. When a candidate refractive index corresponds to multiple over-limit bands, the over-limit band with the largest width is selected as the over-limit band width corresponding to that candidate refractive index. The preset width threshold is determined jointly by the full-spectrum acquisition range, the sampling wavelength interval, and the allowable fitting deviation range.

[0069] Candidate refractive indices whose over-limit band width does not exceed a preset width threshold are identified as candidate refractive indices that satisfy the fitting conditions. When only one candidate refractive index satisfies the fitting conditions, this candidate refractive index is determined as the real-time refractive index of the film under test. When multiple candidate refractive indices satisfy the fitting conditions, the maximum spectral response deviations corresponding to each candidate refractive index are further compared, and the candidate refractive index with the smaller maximum spectral response deviation is determined as the real-time refractive index. The maximum spectral response deviation refers to the maximum value among all sampling wavelength spectral response deviations corresponding to the same candidate refractive index. The real-time refractive index refers to the refractive index of the film under test obtained by fitting the full-spectrum sampling data and monitoring wavelength sampling data within the current modulation period, and is used to subsequently update the film thickness calculation algorithm.

[0070] The preset film thickness calculation algorithm refers to the calculation rules that are pre-called before the deposition of the current film under test and used to describe the correspondence between monitoring wavelength sampling data, full-spectrum sampling data, refractive index parameters, film thickness, and optical thickness. This algorithm can be jointly determined by the material parameters of the film under test, substrate parameters, monitoring wavelength, incident angle parameters, and the calculation relationship of the thin film feature matrix. Specifically, the material type, target optical thickness, and monitoring wavelength of the current film under test are read; the refractive index reference value of the film under test, the refractive index of the substrate, and absorption parameters are read; the incident angle reference value and optical path correction parameters are read; and these parameters are then input into the thin film optical calculation program to generate the correspondence between the monitoring wavelength response and the film thickness, as well as the calculation relationship between the full-spectrum response and the refractive index. This calculation relationship is then used as the preset film thickness calculation algorithm.

[0071] After obtaining the real-time refractive index of the film to be tested, step S4 is executed to update the preset film thickness calculation algorithm online. The real-time refractive index is compared with the refractive index reference value to obtain the refractive index difference between the two. The film thickness monitoring algorithm update parameters refer to the data set used to adjust the original calculation relationship in the preset film thickness calculation algorithm, which includes at least the refractive index offset parameter, the optical thickness conversion correction parameter, and the coating stop judgment correction parameter. Among them, the refractive index offset parameter is obtained by subtracting the refractive index reference value from the real-time refractive index, and is used to indicate the degree of deviation of the current film refractive index from the design or calibration state; the optical thickness conversion correction parameter is determined by the ratio of the real-time refractive index to the refractive index reference value, and is used to correct the conversion relationship between the physical thickness and optical thickness of the film; the coating stop judgment correction parameter is used to adjust the judgment condition corresponding to the current film reaching the target optical film thickness.

[0072] During the update process, the refractive index offset parameter is first written into the refractive index parameter field of the preset film thickness calculation algorithm. This adjusts the film refractive index, which was originally calculated using a refractive index reference value, to the real-time refractive index corresponding to the current modulation period. If the preset film thickness calculation algorithm uses the thin film feature matrix calculation relationship, the real-time refractive index is substituted into the film refractive index field of the thin film feature matrix, and the phase thickness and theoretical spectral response are recalculated based on the updated film refractive index. If the preset film thickness calculation algorithm uses a pre-generated response mapping table, the response mapping table corresponding to the current refractive index is determined according to the refractive index offset parameter, or the original response mapping table is corrected by refractive index offset to obtain the response mapping relationship corresponding to the current modulation period.

[0073] The optical thickness conversion correction parameter is also written into the optical thickness conversion item in the preset film thickness calculation algorithm. The original optical thickness conversion item uses the product relationship between the refractive index reference value and the film thickness as the conversion basis. After being written into the optical thickness conversion correction parameter, the original optical thickness conversion item is corrected according to the proportional relationship between the real-time refractive index and the refractive index reference value, so that the candidate optical thickness corresponding to the same candidate film thickness is recalculated according to the real-time refractive index.

[0074] When determining the target film thickness and the correction parameters for stopping coating, the optical thickness conversion correction parameters are also called. Based on the target optical film thickness and the corrected optical thickness conversion relationship, the target film thickness under the current real-time refractive index conditions is calculated, and then the candidate film thickness that is close to the target film thickness is selected from the candidate film thickness sequence as the target candidate film thickness.

[0075] To further obtain the film thickness estimation range corresponding to the current modulation period, multiple candidate film thicknesses are generated sequentially within this estimation range according to the thickness step size in the algorithm configuration file, forming a candidate film thickness sequence. The thickness step size can be determined based on the control precision of the target optical film thickness, the fluctuation range of the film deposition rate, and computational capabilities.

[0076] For each candidate film thickness in the candidate film thickness sequence, the corresponding candidate optical thickness is calculated using the real-time refractive index. The candidate optical thickness is obtained by multiplying the real-time refractive index by the candidate film thickness. When the influence of the incident angle needs to be considered in the film thickness calculation, the incident angle parameter in the device calibration file is also read, and the cosine correction factor corresponding to the incident angle is introduced into the candidate optical thickness calculation. Each candidate film thickness and its corresponding candidate optical thickness are paired into data pairs, and first mapping data is generated according to the order of the candidate film thicknesses. The first mapping data is used to update the conversion relationship between the refractive index parameter and the optical thickness of the film to be measured.

[0077] Based on the real-time refractive index, monitoring wavelength, and candidate film thickness sequence, the theoretical response data of the monitoring wavelength corresponding to each candidate film thickness is calculated. For each candidate film thickness, the phase thickness is first calculated based on the real-time refractive index, monitoring wavelength, and candidate film thickness. Then, the phase thickness is input into the transmittance calculation relationship, reflectance calculation relationship, or interference response calculation relationship in the preset film thickness calculation algorithm to obtain the theoretical response data corresponding to the candidate film thickness at the monitoring wavelength. The theoretical response data can be theoretical transmittance, theoretical reflectance, or theoretical light intensity response corresponding to the photoelectric acquisition value; the specific type is determined by the current film thickness monitoring mode. Each candidate film thickness and its corresponding monitoring wavelength theoretical response data are paired to generate a second mapping data. The second mapping data is used to update the correspondence between the monitoring wavelength sampling data and the film thickness to be measured.

[0078] After generating the first and second mapping data, the stop-deposition judgment correction parameters corresponding to the current modulation cycle are determined based on the target optical film thickness, real-time refractive index, and candidate film thickness sequence. The target optical film thickness is provided by the film system design document or the current film process document. First, the target film thickness is calculated based on the target optical film thickness and real-time refractive index. Then, the candidate film thickness with the smallest difference from the target film thickness is selected from the candidate film thickness sequence as the target candidate film thickness. Subsequently, the theoretical response data of the monitoring wavelength corresponding to the target candidate film thickness is read from the second mapping data, and this data is used as the target stop point data under the current modulation cycle. The difference between the target film thickness and the target candidate film thickness is also calculated to obtain the target thickness discrete deviation. The preset stop-deposition allowable deviation is adjusted based on the target thickness discrete deviation to obtain the corrected stop-deposition allowable deviation. The preset stop-deposition allowable deviation is determined by the film thickness control requirements in the film system design document and the historical deposition calibration results. The target stop point data and the corrected stop-deposition allowable deviation together constitute the stop-deposition judgment correction parameters.

[0079] The film thickness calculation mapping relationship in the preset film thickness calculation algorithm is updated online using parameters updated by the film thickness monitoring algorithm, first mapping data, second mapping data, and stop-deposition judgment correction parameters. Online updating means that instead of re-establishing a new film thickness calculation algorithm during the current film deposition process, the refractive index reference parameter in the preset film thickness calculation algorithm is replaced with the parameter corresponding to the real-time refractive index; the original optical thickness conversion relationship is replaced with the first mapping data; the original correspondence between the monitoring wavelength sampling data and the film thickness is replaced with the second mapping data; and the stop point data and allowable deviation in the original stop-deposition judgment conditions are replaced with stop-deposition judgment correction parameters. After the update, an updated film thickness calculation algorithm is formed. This updated film thickness calculation algorithm retains the basic calculation framework of the preset film thickness calculation algorithm, while adjusting the refractive index parameter, thickness conversion relationship, monitoring wavelength response correspondence, and stop-deposition judgment conditions according to the real-time refractive index of the current modulation cycle.

[0080] After determining the monitoring wavelength sampling data corresponding to the current modulation period, the signal is matched with the second mapping data: the monitoring wavelength sampling data of the current modulation period is compared with the theoretical response data of the monitoring wavelength corresponding to each candidate film thickness in the second mapping data, and the candidate film thickness with the smallest absolute difference is taken as the actual thickness of the film to be tested corresponding to the current modulation period. When there are multiple candidate film thicknesses with the same absolute difference or within the same minimum difference range, the actual thickness of the film to be tested in the previous modulation period is read, and the candidate film thickness with the same change relationship as the actual thickness in the previous modulation period is selected. If the current modulation period is the first effective modulation period of the current film to be tested, the candidate film thickness that is closer to the estimated film thickness in step S3 is selected.

[0081] After obtaining the actual thickness of the film to be measured, the real-time optical film thickness is calculated based on the real-time refractive index and the actual thickness of the film to be measured. Specifically, the real-time refractive index is multiplied by the actual thickness of the film to be measured to obtain the real-time optical film thickness. When the current film thickness calculation mode needs to consider the influence of the incident angle, the incident angle parameter is read from the calibration file, and the cosine correction factor corresponding to the incident angle is introduced into the calculation. The incident angle parameter is obtained through optical path calibration and stored in the calibration file.

[0082] The target optical film thickness is provided by the film system design document. Before the current film layer under test begins deposition, the current film layer task parameters are read and written. The real-time optical film thickness is compared with the target optical film thickness, and the difference between the two is calculated to obtain the optical thickness deviation of the film layer under test. The optical thickness deviation of the film layer under test is used to indicate how close the current film layer optical thickness is to the target optical film thickness. Based on the real-time optical film thickness, the target optical film thickness, and the optical thickness deviation of the film layer under test, the film thickness monitoring result corresponding to the current modulation cycle is generated. The film thickness monitoring result includes the current modulation cycle number, real-time refractive index, actual thickness of the film layer under test, real-time optical film thickness, target optical film thickness, optical thickness deviation of the film layer under test, and the current stop deposition judgment status.

[0083] The allowable deviation for stopping the coating process corresponding to the current modulation cycle is determined based on the stop-coating judgment correction parameters. The allowable deviation refers to the allowable deviation range for determining whether the real-time optical film thickness has approached the target optical film thickness. After reading this allowable deviation, the absolute value of the optical thickness deviation of the film to be measured is compared with the allowable deviation. When the absolute value of the optical thickness deviation of the film to be measured is not greater than the allowable deviation, it is determined that the current film thickness monitoring result meets the preset stop-coating condition, and a stop-coating signal is output to the coating equipment control unit.

[0084] When the absolute value of the optical thickness deviation of the film under test exceeds the allowable deviation for stopping deposition, it is determined that the current modulation cycle has not directly met the deposition-stopping condition, and it is further determined whether the next modulation cycle may exceed the target optical film thickness. Specifically, the real-time optical film thickness of the previous modulation cycle is read from the data buffer of the current film layer, and the difference between the real-time optical film thickness of the current modulation cycle and the real-time optical film thickness of the previous modulation cycle is calculated to obtain the real-time optical film thickness change. If the current modulation cycle is the first valid modulation cycle of the film under test, the prediction judgment based on the previous modulation cycle is not performed, but the film thickness monitoring of the next modulation cycle continues. If data from the previous modulation cycle exists, the real-time optical film thickness of the current modulation cycle is added to the real-time optical film thickness change to obtain the predicted optical film thickness for the next modulation cycle.

[0085] The prediction results indicate a tendency for the next modulation cycle to exceed the target optical film thickness. This means that the real-time optical film thickness of the current modulation cycle has not yet reached the stop-deposition range corresponding to the target optical film thickness, but the predicted optical film thickness of the next modulation cycle has already exceeded the target optical film thickness, or the direction of the difference between the predicted and target optical film thicknesses has changed relative to the current modulation cycle. Whether this tendency has occurred is determined by comparing the signs of the optical thickness deviation of the current modulation cycle and the predicted optical thickness deviation of the next modulation cycle. When the signs are different, or the absolute value of the predicted optical thickness deviation is less than the allowable stop-deposition deviation, it is determined that the next modulation cycle has a tendency to exceed the target optical film thickness.

[0086] When a tendency to exceed the allowable limit exists, a compensation judgment is made based on the stop-deposition judgment correction parameters to compensate for the preset stop-deposition conditions. The compensation judgment refers to determining whether a stop-deposition signal needs to be output prematurely, considering the proximity between the predicted optical film thickness and the target optical film thickness in the next modulation cycle, even if the current modulation cycle does not directly meet the allowable stop-deposition deviation. Specifically, the absolute value of the difference between the predicted optical film thickness and the target optical film thickness in the next modulation cycle is calculated to obtain the prediction deviation; then, the corrected allowable stop-deposition deviation and the target stop point data are read from the stop-deposition judgment correction parameters; when the prediction deviation is not greater than the corrected allowable stop-deposition deviation, and the difference between the monitoring wavelength sampling data corresponding to the current modulation cycle and the target stop point data is within the preset stop point proximity range, the compensation judgment is deemed satisfied, and a stop-deposition signal is output. The preset stop point proximity range is determined by the repeated acquisition fluctuation range of the monitoring wavelength sampling data and the historical stop-deposition calibration results.

[0087] If the compensation judgment is not met, it means that although the current modulation cycle predicts that the next modulation cycle may approach or exceed the target optical film thickness, the proximity between the current monitoring wavelength sampling data and the target stop point data is insufficient, or the prediction deviation still exceeds the corrected allowable deviation for stopping the deposition. In this case, no stop deposition signal is output. Instead, film thickness monitoring for the next modulation cycle continues, and the real-time optical film thickness, the actual thickness of the film under test, the real-time refractive index, and the film thickness monitoring results of the current modulation cycle are written into the data buffer of the current film layer as historical data for the prediction judgment of the next modulation cycle.

[0088] If the prediction results indicate that there is no trend of exceeding the target optical film thickness in the next modulation cycle, it is determined that the film thickness monitoring result of the current modulation cycle does not meet the preset stop-deposition condition, and film thickness monitoring continues for the next modulation cycle. After the start of the next modulation cycle, the processing is repeated in the order of periodic modulation, synchronous sampling, refractive index fitting, film thickness calculation algorithm update, and stop-deposition judgment until the film thickness monitoring result meets the preset stop-deposition condition and a stop-deposition signal is output.

[0089] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the flow or function according to the embodiments of the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0090] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0091] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0092] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0093] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0094] In the embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0095] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0096] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0097] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0098] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A film thickness monitoring method based on optical signal modulation and fiber coupling, characterized in that, Includes the following steps: S1, periodically modulate the detection optical parameters and light transmission state during the thin film deposition process, transmit them to the film layer under test after optical signal coupling processing, collect the return optical signal after the film layer under test is applied, and allocate them to the first acquisition channel and the second acquisition channel respectively. The first acquisition channel is used to collect the monitoring wavelength optical signal, and the second acquisition channel is used to collect the full spectrum optical signal. S2, based on the modulation period corresponding to the detection optical parameters, the monitoring wavelength optical signal and the full spectrum optical signal are synchronously segmented to obtain the monitoring wavelength sampling data and full spectrum sampling data within the same modulation period, so as to establish the synchronous sampling data corresponding to the monitoring of the thickness of the film to be tested. S3, based on the synchronous sampling data, extract the spectral response variation characteristics of the film under test at different wavelengths, and fit the refractive index of the film under test to obtain the real-time refractive index of the film under test; S4. Generate corresponding film thickness monitoring algorithm update parameters according to the real-time refractive index, and update the film thickness calculation mapping relationship in the preset film thickness calculation algorithm online to obtain the updated film thickness calculation algorithm. S5, input the real-time refractive index and synchronous sampling data into the updated film thickness calculation algorithm to calculate the real-time optical film thickness of the film to be tested, and generate film thickness monitoring results based on the comparison results between the real-time optical film thickness and its corresponding target optical film thickness. When the film thickness monitoring results meet the preset stop-deposition conditions, output a stop-deposition signal. The S2 includes: Acquire the monitoring wavelength optical signal output from the first acquisition channel and the full spectrum optical signal output from the second acquisition channel, and read their respective sampling times; Based on the period identifier, the period assignment of the monitoring wavelength optical signal and the full spectrum optical signal is determined to identify the period number and time period corresponding to each sampled data. The time period includes the film detection time period, the reference sampling time period and the background sampling time period. The monitoring wavelength optical signal and the full spectrum optical signal belonging to the same period number are taken as the optical signal to be segmented within the same modulation period. According to the time period boundary corresponding to the period identifier, the optical signal to be segmented is synchronously intercepted and the time period is marked to obtain the monitoring wavelength segmented data and the full spectrum segmented data. The stable sampling area is extracted from the monitoring wavelength segmented data and the full spectrum segmented data. Transitional sampling points caused by the switching of detection light state in each time period are removed, and the representative light intensity value corresponding to each time period is determined based on the sampling value in the stable sampling area. The light intensity representative value is used to characterize the stable light intensity values ​​of the return light signal, the reference light signal, and the background light signal within the same modulation period in the corresponding time period; The S2 further includes: Based on the light intensity representative values ​​corresponding to the film detection time period, the reference sampling time period and the background sampling time period within the same modulation period, the monitoring wavelength segmented data and the full spectrum segmented data are respectively subjected to background subtraction and reference normalization processing to obtain the monitoring wavelength sampling data and the initial full spectrum sampling data. Based on the correspondence between the monitoring wavelength sampling data and the initial full-spectrum sampling data at the monitoring wavelength, the channel response correction parameter corresponding to the same modulation period is determined, and the initial full-spectrum sampling data is corrected based on the channel response correction parameter to obtain the full-spectrum sampling data. By associating the monitoring wavelength sampling data, full-spectrum sampling data, and corresponding period identifiers within the same modulation period, synchronous sampling data corresponding to the monitoring of the thickness of the film to be measured is established.

2. The film thickness monitoring method based on optical signal modulation and optical fiber coupling as described in claim 1, characterized in that, The periodic modulation includes: Obtain the film thickness monitoring parameters corresponding to the film layer to be tested. The film thickness monitoring parameters include at least the monitoring wavelength, sampling frequency, and film deposition rate. The modulation period of the detection light parameter is determined based on the film thickness monitoring parameters, so that the change in the thickness of the film layer under test within a single modulation period is within the preset allowable range of film thickness monitoring. The modulation period is divided into a film detection time period, a reference sampling time period, and a background sampling time period, and then combined to form the modulation period for detecting optical parameters. The film detection time period is used to acquire the returned light signal after the film layer under test is applied, the reference sampling time period is used to acquire the reference light signal, and the background sampling time period is used to acquire the background light signal. The return optical signal, reference optical signal and background optical signal acquired within the same modulation period are bound to the corresponding period identifiers. The period identifiers are used to determine the temporal correspondence of each sampled data within the same modulation period.

3. The film thickness monitoring method based on optical signal modulation and optical fiber coupling as described in claim 2, characterized in that, The specific process of assigning data to the first acquisition channel and the second acquisition channel is as follows: Based on the period identifier, the return optical signal, reference optical signal and background optical signal formed within the same modulation period are taken as optical signals to be allocated. According to the monitoring wavelength acquisition requirements of the first acquisition channel and the full spectrum acquisition requirements of the second acquisition channel, the first channel allocation coefficient corresponding to the first acquisition channel and the second channel allocation coefficient corresponding to the second acquisition channel are determined. According to the first channel allocation coefficient and the second channel allocation coefficient, the optical signal to be allocated is optically path allocated so that the corresponding optical signal enters the first acquisition channel and the second acquisition channel respectively. In the first acquisition channel, the light signal entering the channel is monitored by selecting the wavelength to obtain the monitoring wavelength light signal. In the second acquisition channel, the light signal entering the channel is acquired in the full spectrum range to obtain the full spectrum light signal. According to the temporal correspondence of each time period in the period identifier, the monitoring wavelength optical signal and the full spectrum optical signal are marked with time periods respectively. The acquisition results located in the film layer detection time period are marked as return light acquisition data, the acquisition results located in the reference sampling time period are marked as reference light acquisition data, and the acquisition results located in the background sampling time period are marked as background light acquisition data. By correlating the acquisition results of the first acquisition channel and the second acquisition channel, which have the same period identifier, we can obtain the monitoring wavelength optical signal and the full spectrum optical signal under the same modulation period reference.

4. The film thickness monitoring method based on optical signal modulation and optical fiber coupling as described in claim 1, characterized in that, The S3 includes: Extract the full-spectrum sampling data, monitoring wavelength sampling data, and corresponding period identifiers within the same modulation period from the synchronous sampling data; According to the full-spectrum acquisition range corresponding to the second acquisition channel, the full-spectrum sampling data is arranged in wavelength order to obtain the spectral response sequence corresponding to the current modulation period; The spectral response sequence corresponding to the current modulation period is compared with the initial spectral response sequence corresponding to the start of the deposition of the film under test, wavelength by wavelength, to obtain the spectral response change at different wavelengths, and the corresponding spectral response change trend is determined based on the difference in spectral response change between adjacent wavelengths.

5. The film thickness monitoring method based on optical signal modulation and optical fiber coupling as described in claim 4, characterized in that, The S3 further includes: Based on the spectral response sequence, the spectral response variation at different wavelengths, the spectral response variation trend, and the monitoring wavelength sampling data, the spectral response variation characteristics of the film under test at different wavelengths are generated. Based on the refractive index reference value corresponding to the film layer to be tested, a candidate range of refractive index is determined, and a candidate refractive index set is generated within the candidate range of refractive index. Based on the monitoring wavelength sampling data and the deposition process parameters of the film to be tested, the estimated range of film thickness corresponding to the current modulation period is determined, and the estimated value of film thickness corresponding to the current modulation period is determined based on the preset film thickness calculation algorithm. Each candidate refractive index and the estimated film thickness are input into a preset film thickness calculation algorithm to obtain the theoretical spectral response data corresponding to each candidate refractive index. The theoretical spectral response data is then compared with the spectral response sequence corresponding to the current modulation period to obtain the spectral response deviation corresponding to each candidate refractive index. Based on the spectral response changes at different wavelengths, sampling wavelengths with spectral response deviations greater than a preset deviation threshold are marked as deviation-exceeding sampling wavelengths. The width of the excess band is determined based on the deviation-exceeding sampling wavelengths, and the candidate refractive index with the excess band width not exceeding a preset width threshold is determined as the real-time refractive index of the film to be tested.

6. The film thickness monitoring method based on optical signal modulation and optical fiber coupling as described in claim 5, characterized in that, The S4 includes: Based on the difference between the real-time refractive index and the refractive index reference value, film thickness monitoring algorithm update parameters are generated. The film thickness monitoring algorithm update parameters include at least refractive index offset parameters, optical thickness conversion correction parameters, and coating stop determination correction parameters. Based on the real-time refractive index and the film thickness estimation range corresponding to the current modulation period, a candidate film thickness sequence is constructed, and the candidate optical thickness corresponding to each candidate film thickness is calculated to generate the first mapping data between the candidate thickness and the candidate optical thickness. Based on the real-time refractive index, monitoring wavelength and candidate film thickness sequence, calculate the theoretical response data of the monitoring wavelength corresponding to each candidate film thickness, and generate a second mapping data between the candidate thickness and the theoretical response data of the monitoring wavelength. Based on the target optical film thickness, the real-time refractive index, and the candidate film thickness sequence, determine the stop-deposition judgment correction parameters corresponding to the current modulation cycle; The membrane thickness monitoring algorithm is used to update the parameters, the first mapping data and the second mapping data, and the membrane thickness calculation mapping relationship in the preset membrane thickness calculation algorithm is updated online to obtain the updated membrane thickness calculation algorithm. The film thickness calculation mapping relationship includes at least: the conversion relationship between the refractive index parameter and the optical thickness of the film to be measured, the correspondence between the monitoring wavelength sampling data and the thickness of the film to be measured, and the correlation between the optical thickness deviation of the film to be measured and the stop-deposition judgment condition.

7. The film thickness monitoring method based on optical signal modulation and optical fiber coupling as described in claim 6, characterized in that, The S5 includes: The real-time refractive index and synchronous sampling data are input into the updated film thickness calculation algorithm; Based on the updated film thickness calculation algorithm, the monitoring wavelength sampling data corresponding to the current modulation period is matched with the second mapping data to determine the actual thickness of the film layer to be measured corresponding to the current modulation period. Based on the real-time refractive index and the actual thickness of the film to be tested, the real-time optical thickness of the film to be tested is calculated, and the real-time optical thickness is compared with the target optical thickness to obtain the optical thickness deviation of the film to be tested. Based on the real-time optical film thickness, the target optical film thickness, and the optical thickness deviation of the film layer under test, the film thickness monitoring result corresponding to the current modulation period is generated.

8. The film thickness monitoring method based on optical signal modulation and optical fiber coupling as described in claim 7, characterized in that, The S5 also includes: Based on the stop-deposition determination correction parameters, the allowable deviation for stop-deposition corresponding to the current modulation cycle is determined. When the optical thickness deviation of the film to be measured meets the allowable deviation for stop-deposition, it is determined that the film thickness monitoring result meets the preset stop-deposition condition, and a stop-deposition signal is output. When the optical thickness deviation of the film to be tested does not meet the allowable deviation for stopping the coating, the optical thickness of the next modulation cycle is predicted based on the real-time optical thickness change corresponding to the current modulation cycle and the previous modulation cycle. If the prediction results indicate that the next modulation cycle tends to exceed the target optical film thickness, then the preset stop-deposition condition is compensated and determined based on the stop-deposition determination correction parameter, and a stop-deposition signal is output when the compensation determination is satisfied. If the compensation determination is not met, the film thickness monitoring will continue in the next modulation cycle. If the prediction results indicate that there is no tendency to exceed the target optical film thickness in the next modulation cycle, then it is determined that the film thickness monitoring result of the current modulation cycle does not meet the preset stop-deposition condition, and the film thickness monitoring of the next modulation cycle continues.

Images