System and method for improving film thickness measurement precision through reflection spectrometry

Abstract

The invention discloses a system for improving the precision of film thickness measurement by a reflection spectrometry, which is characterized by comprising a white light source for emitting light in a wide spectral range; the main spectrograph is used for measuring the reflection spectrum of the measured object; the monitoring spectrometer is used for monitoring the spectral intensity of the light source in real time; the optical fiber assembly comprises a multi-core optical fiber and a single-core optical fiber and is used for conducting light from the light source to a measured object and conducting reflected light to the main spectrograph; and the data processing unit is used for calibrating the incident light intensity in real time according to the data of the main spectrograph and the monitoring spectrograph and calculating the thickness of the thin film. According to the system for improving the precision of measuring the thickness of the thin film through the reflection spectrum method, the repeatability precision of measurement and the stability of long-term measurement are improved.

Description

Technical Field

[0001] This invention relates to the field of optical thin film measurement technology, and in particular to a system and method for improving the accuracy of thin film thickness measurement by reflectance spectroscopy, which can be applied to thin film quality inspection and process control in fields such as semiconductor manufacturing, optical coating, display panels, and photovoltaic materials. Background Technology

[0002] Reflectance spectroscopy is a commonly used technique for measuring the thickness of transparent and semi-transparent thin films, and it is widely applied in semiconductors, optical coatings, liquid crystal displays, photovoltaics, and other fields. This method is based on the principle of white light interference. By analyzing the interference spectrum formed by the reflected light from the thin film surface and the substrate interface, the thickness of the thin film and its optical constants (refractive index n and extinction coefficient k) can be deduced.

[0003] A typical reflectance spectroscopy film thickness measurement system includes a broadband white light source, an optical fiber transmission component, a spectrometer, and a data processing unit. Light emitted from the source is transmitted to the sample surface via the optical fiber, and the reflected light is transmitted to the spectrometer via a collecting optical fiber to acquire reflectance spectral data. The film thickness can be calculated by fitting the measured reflectance spectrum with a theoretical reflectance spectrum based on an optical model.

[0004] In this measurement process, the stability of the incident light intensity on the sample surface is a key factor affecting the repeatability and accuracy of the measurement. In actual measurements, the incident light intensity is easily affected by factors such as fluctuations in the light source output, changes in fiber coupling efficiency, and variations in the distance between the probe and the sample, making it impossible to obtain directly and accurately. To eliminate this influence, existing technologies typically use standard samples (such as silicon wafers with known optical constants and thicknesses) for calibration: first, the reflection spectrum of the standard sample is measured, and the incident light intensity A0 is deduced from its known optical parameters. In subsequent measurements, this is used as a constant and substituted into the model to calculate the thickness of the film to be measured.

[0005] However, this method has obvious limitations: First, the positions of the standard sample and the sample to be tested are difficult to be completely consistent. Especially in high-precision automatic measurement, even small positional differences can introduce significant calibration errors.

[0006] Secondly, frequent calibration using standard samples will consume measurement time, reduce equipment throughput, and affect production efficiency.

[0007] Furthermore, this method cannot overcome the noise caused by the instantaneous fluctuations of the light source itself, which limits further improvement in measurement repeatability accuracy.

[0008] In addition, in pursuit of stability, existing systems often require the use of highly stable light sources, which leads to a significant increase in equipment costs.

[0009] Therefore, how to effectively suppress the impact of light source fluctuations and improve the long-term repeatability and instantaneous stability of film thickness measurement by reflectance spectroscopy without significantly increasing costs remains a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0010] The present invention aims to provide a system and method for improving the accuracy of thin film thickness measurement by reflectance spectroscopy, thereby improving the repeatability accuracy of the measurement, ensuring the stability of long-term measurement, and increasing measurement efficiency and productivity.

[0011] To achieve the above objectives, in a first aspect, the present invention provides a system for improving the accuracy of thin film thickness measurement by reflectance spectroscopy, comprising: White light sources are used to emit light across a wide spectrum. The main spectrometer is used to measure the intensity of the reflected spectrum of the object being measured. A monitoring spectrometer is used to monitor the spectral intensity of a light source in real time. The optical fiber assembly includes a multi-core optical fiber and a single-core optical fiber. The single-core optical fiber is used to transmit part of the spectral signal of the light source to the monitoring spectrometer, and the multi-core optical fiber is used to transmit light from the light source to the object under test and to transmit the reflected light to the main spectrometer. The data processing unit is used to calibrate the incident light intensity in real time based on data from the main spectrometer and the monitoring spectrometer, and to calculate the film thickness.

[0012] By adopting the above technical solution, a collaborative measurement system of "white light source - dual spectrometer - fiber optic assembly - data processing unit" was constructed: the main spectrometer acquires the core data of the reflectance spectrum of the object under test, the monitoring spectrometer captures the changes in light source intensity in real time, and the data processing unit dynamically calibrates the incident light intensity based on the data of the dual spectrometer. From the hardware architecture perspective, it breaks through the limitation of traditional single spectrometers that rely on fixed incident light parameters, and can simultaneously solve the impact of instantaneous fluctuations and long-term attenuation of the light source on the measurement accuracy, laying a hardware foundation for improving the repeatability and long-term stability of thin film thickness measurement.

[0013] Furthermore, the multi-core optical fiber transmits light to the film being measured, and the light reflected back from the film passes through one of the single cores of the multi-core optical fiber to transmit the reflected spectral signal to the main spectrometer.

[0014] By adopting the above technical solution, precise splitting of optical path functions is achieved: single-core optical fiber focuses on light source intensity monitoring, directly extracting light from the light source to ensure that the monitoring signal accurately reflects the original state of the light source; multi-core optical fiber is used for the measurement optical path, which can uniformly conduct the light source to the surface of the object being measured, reducing the reflection spectrum deviation caused by uneven illumination. The two work independently and collaboratively, avoiding interference between the monitoring optical path and the measurement optical path, while ensuring the synchronization of the light source monitoring and measurement processes, thus improving the stability of optical path transmission and the accuracy of data acquisition.

[0015] Furthermore, the monitoring spectrometer has the same configuration and chip as the main spectrometer to ensure consistent signal acquisition.

[0016] By adopting the above technical solution, the calibration deviation caused by the difference in hardware performance between the two spectrometers is eliminated: the same configuration (such as the same optical lens and signal processing module) and chip can ensure that the signal acquisition efficiency, wavelength resolution and response speed of the two are completely consistent, avoiding the "asynchronous" deviation between the light intensity signal of the monitoring spectrometer and the reflection spectrum signal of the main spectrometer due to different hardware parameters, reducing the error when the data processing unit calculates the calibration coefficient, and ensuring the reliability of the calibration logic.

[0017] Furthermore, the data processing unit calculates the calibration coefficient Kt using the following formula: Kt = Bt / B0 × K0; Where Bt is the light intensity measured by the monitoring spectrometer at time t, B0 is the reflection spectral intensity of the standard thickness silicon wafer, and K0 is the initial calibration coefficient.

[0018] By adopting the above technical solution, a dynamic calibration mechanism is established: Bt / B0 can quantify the intensity change of the light source at time t relative to the initial calibration state (such as instantaneous fluctuations or long-term attenuation). Combined with the initial calibration coefficient K0, the calibration coefficient Kt adapted to the light source state at time t can be quickly calculated. This formula directly correlates the light source intensity change with the calibration coefficient, realizing the real-time updating of the calibration coefficient with the light source state, avoiding the shortcomings of traditional fixed calibration parameters that cannot adapt to the dynamic changes of the light source.

[0019] Furthermore, the initial calibration coefficient K0 is obtained by the following formula: K0 = B0 / A0; Where A0 is the incident light intensity obtained through standard silicon wafer calibration.

[0020] By adopting the above technical solution, an initial reference correlation is established: the optical constants and thickness of the standard silicon wafer are known, allowing for the accurate calculation of the initial incident light intensity A0; B0 is the light source intensity measured by the monitoring spectrometer at the initial calibration moment; K0=B0 / A0 quantifies the ratio between the monitoring spectrometer signal and the actual incident light intensity, providing a core reference for subsequent dynamic calibration. This method requires only one standard silicon wafer calibration to establish the reference, eliminating the need for repeated calibrations, thus ensuring reference accuracy while simplifying the operation process.

[0021] Secondly, this application provides a method for improving the accuracy of thin film thickness measurement by reflectance spectroscopy, comprising the following steps: Initial calibration was performed using a standard silicon wafer to obtain the initial incident light intensity A0 and the monitoring light intensity B0; In actual measurements, the light intensity Bt measured by the monitoring spectrometer is acquired in real time. Calculate the real-time calibration coefficient Kt based on Bt and B0; The incident light intensity is calibrated in real time using Kt; The film thickness is calculated based on the calibrated incident light intensity.

[0022] By adopting the above technical solution, a complete measurement process of "initial calibration - real-time calibration - thickness calculation" is formed: the initial calibration only needs to be performed once to establish the reference parameters (A0, B0). In subsequent actual measurements, the incident light intensity is dynamically calibrated by acquiring Bt in real time and calculating Kt. Finally, the thickness is calculated based on the calibrated parameters. This process avoids the traditional method of repeatedly calibrating the standard silicon wafer before each measurement, eliminating errors caused by differences in calibration positions and saving calibration time. At the same time, real-time calibration solves the problem of dynamic changes in the light source, balancing accuracy and efficiency.

[0023] Furthermore, the formula for calculating the real-time calibration coefficient Kt is as follows: Kt = (Bt / B0) × K0.

[0024] By adopting the above technical solution, the dynamic adaptability of the calibration coefficient is ensured: Bt reflects the actual intensity of the light source at time t, B0 is the initial reference intensity, and the ratio of the two directly reflects the magnitude of the light source intensity change; after multiplying by the initial calibration coefficient K0, Kt can accurately correlate "monitoring light intensity change" with "incident light intensity calibration requirements", so that the calibrated incident light intensity parameter (At=Kt×A0) completely matches the light source state at time t, effectively compensating for the measurement deviation caused by instantaneous fluctuations and long-term attenuation of the light source.

[0025] Furthermore, the calculation of the film thickness is based on the following model: R / (A0×Kt)=f (N,K,THK), Where R is the reflectance spectral intensity, f(N,K,THK) is the corresponding reflectance fitting function of the measured film, N and K are optical constants, and THK is the film thickness.

[0026] By adopting the above technical solution, dynamic calibration is integrated into the core model of thickness calculation: A0×Kt represents the incident light intensity At after calibration at time t, and the ratio of R to At is the true reflectivity. This reflectivity is precisely matched with the functional relationship f(N,K,THK) between the thin film optical constants (N,K) and thickness (THK). Compared with the traditional model using a fixed A0, this model can correct the incident light intensity deviation in real time, ensuring the accuracy of reflectivity calculation, thereby improving the accuracy and repeatability of the thin film thickness fitting results.

[0027] Furthermore, the data acquisition time of the main spectrometer and the monitoring spectrometer is kept consistent.

[0028] By adopting the above technical solution, the calibration deviation caused by time difference is eliminated: the main spectrometer collects the reflection spectrum (R) and the monitoring spectrometer collects the light source intensity (Bt) synchronously, which can ensure that Bt reflects the light source intensity of "the incident light corresponding to R", avoids the inability of Bt to match the light source state corresponding to R due to asynchronous collection time (such as the instantaneous fluctuation of the light source during the interval between two collections), ensures the temporal correlation between calibration data and measurement data, and improves the effectiveness of dynamic calibration.

[0029] Furthermore, the real-time acquisition of the light intensity Bt measured by the monitoring spectrometer is synchronized with the acquisition of the reflectance spectrum signal by the main spectrometer, specifically including: The data processing unit sends a synchronization trigger signal to the main spectrometer and the monitoring spectrometer; The main spectrometer and the monitoring spectrometer synchronously acquire a frame of reflectance spectrum signal and a frame of light source monitoring signal, respectively, based on the synchronization trigger signal, to obtain R and Bt at that moment.

[0030] By adopting the above technical solution, high-precision synchronization of data acquisition by the two spectrometers is achieved: the synchronization trigger signal sent by the data processing unit provides a unified time reference for the two spectrometers, ensuring that both start data acquisition at the same instant, avoiding the acquisition time difference caused by the difference in the response delay of the spectrometers themselves. This method can control the time synchronization error of the two spectrometers to the microsecond level, which is especially suitable for scenarios with millisecond-level instantaneous fluctuations in the light source, further improving the accuracy of dynamic calibration and ensuring the repeatability of thin film thickness measurement.

[0031] In summary, this application has at least one of the following beneficial technical effects: 1. By using a dual-spectrometer architecture and a dynamic calibration coefficient model, the problems of instantaneous fluctuations and long-term attenuation of the light source are solved simultaneously, significantly improving the repeatability accuracy of thin film thickness measurement and reducing measurement errors after long-term use; 2. Only one initial calibration of the standard silicon wafer is required, eliminating the need for repeated calibration steps before each measurement in traditional solutions, avoiding errors due to differences in calibration positions, saving calibration time, and improving measurement efficiency and productivity; 3. The dual spectrometers use the same configuration and chip, combined with a synchronous trigger acquisition design, to eliminate calibration deviations caused by hardware differences and time differences, and to ensure the reliability and accuracy of dynamic calibration logic; 4. No need to rely on highly stable special light sources; ordinary white light sources can meet the high-precision measurement requirements through dynamic calibration, which greatly reduces hardware procurement and maintenance costs and improves the cost-effectiveness of the equipment. 5. The measurement process is simplified and adapted to automated scenarios. No manual intervention is required after the initial calibration. It can cover the measurement of various thin film types such as dielectrics, semiconductors, and thin film filters, enhancing the applicability and practicality of the equipment. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0033] Figure 1 This is a schematic diagram of light source reflection; Figure 2 This is a schematic diagram of the system for improving the accuracy of thin film thickness measurement by reflectance spectroscopy in this application.

[0034] Figure label: 1. Substrate; 2. Thin film; 3. Lens; 4. Main spectrometer; 5. Data processing unit; 6. Light source; 7. Monitoring spectrometer; 8. Single-core optical fiber; 9. Multi-core optical fiber; 91. Incident light transmission core; 92. Reflected light transmission core. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0037] In the description of this application, it should be understood that the terms "upper", "lower", "left", "right", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0038] The technical solutions of this application will be described in detail below with reference to the accompanying drawings and specific embodiments. Unless otherwise specified, the features in the following embodiments can be combined with each other.

[0039] Example 1: A system for improving the accuracy of thin film thickness measurement using reflectance spectroscopy Please see Figure 1 The instrument analyzes the coherent reflection spectrum of the light reflected from the interface between the surface of the film 2 and the substrate 1. By comparing the reflection spectrum after reflection from the film 2 with the spectral intensity incident on the surface of the film 2, the reflectivity data after passing through the film 2 can be obtained. Given the optical constants of the film 2, the thickness of the film 2 can be obtained by fitting the reflectivity data using an algorithm. In this measurement process, the stability of the incident light is the main factor determining the measurement stability and repeatability.

[0040] Please see Figure 2 This embodiment discloses a system for improving the accuracy of thin film 2 thickness measurement by reflection spectroscopy. It is designed to solve the problems of low accuracy and poor efficiency caused by instantaneous fluctuations and long-term attenuation of the light source in traditional reflection spectroscopy thickness measurement. The system consists of a white light source 6, a main spectrometer 4, a monitoring spectrometer 7, an optical fiber assembly and a data processing unit 5. The components are precisely coordinated through optical and electrical paths to form a complete measurement link.

[0041] The white light source 6 emits light across a wide spectral range, covering 190–2300 nm. It is suitable for various thin-film measurement needs, including dielectrics, semiconductors, and thin-film filters. Its light intensity stability in the uncalibrated state is ≤±2% / h. Employing a wind-cooled design, it can operate continuously for over 8000 hours without relying on expensive special stable light sources; its own fluctuations and attenuation can be compensated for through subsequent dynamic calibration. The main spectrometer 4 measures the reflectance spectrum of the object under test, obtaining the reflectance intensity R. It uses a 2048-pixel linear CCD detection chip with a wavelength resolution <0.4 nm, a spectral acquisition speed ≥10 frames / s, and an adjustable integration time of 1ms–10s. Data is transmitted via a USB 3.0 interface with a latency <10ms. The monitoring spectrometer 7 is used to monitor the spectral intensity of the light source 6 in real time and obtain the light intensity of the light source 6 at different times (B0 in the initial calibration stage and Bt in the actual measurement stage). Its hardware configuration is completely consistent with that of the main spectrometer 4, including the same model of CCD chip, optical lens group and signal processing module, to ensure that the signal acquisition efficiency and wavelength response characteristics of the two are consistent and to avoid calibration deviation caused by hardware differences.

[0042] The fiber optic assembly comprises a multi-core fiber 9 and a single-core fiber 8, both made of quartz with a numerical aperture of 0.22 and a core diameter of 50 μm. The outer sheath is made of wear-resistant polyimide. One end of the single-core fiber 8 is connected to the branch light output port of the white light source 6 (splitting 5%–10% of the light signal from the total output of the light source 6 without affecting the measurement optical path), and the other end is connected to the light input port of the monitoring spectrometer 7, used only to transmit a portion of the spectral signal from the light source 6 to the monitoring spectrometer 7. The multi-core fiber 9 specifically consists of 6 fibers, of which 5 are incident light guiding cores 91, arranged in a ring array. One end is connected to the main output port of the white light source 6, and the other end is aligned with the measurement area of ​​the object under test, used to uniformly transmit most of the light from the light source 6 to the surface under test, forming a uniform light spot with an adjustable diameter of 2–5 mm. The remaining core is a reflected light guiding core 92, located at the center of the ring array. One end is aligned with the reflected light area of ​​the object under test, and the other end is connected to the light input port of the main spectrometer 4, responsible for transmitting the reflected spectral signal to the main spectrometer 4.

[0043] The data processing unit 5 uses an industrial-grade computer equipped with a dedicated data acquisition card and a software system developed based on LabVIEW. It comprises four main modules: data acquisition, calibration calculation, thickness fitting, and result storage. It receives measurement data from the main spectrometer 4 and the monitoring spectrometer 7 via a USB 3.0 data cable and can also send TTL synchronization trigger signals to both spectrometers via GPIO interfaces, ensuring that the acquisition time difference between them is ≤1μs. In terms of data calculation, the data processing unit 5 calculates the real-time calibration coefficient Kt using the formula Kt=(Bt / B0)×K0 (where Bt is the monitored light intensity at time t, B0 is the initial calibration light intensity, and K0 is the initial calibration coefficient). The initial calibration coefficient K0 is obtained by K0=B0 / A0, where A0 is the initial incident light intensity obtained using standard silicon wafer calibration.

[0044] The system's workflow consists of two stages: initial calibration and actual measurement. The initial calibration stage only requires one step: a standard silicon wafer (N-type single crystal silicon, thickness 500μm±1μm, optical constants known) is placed at the stage positioning position, and the stage height is adjusted so that the optical fiber output axis of lens 3 is perpendicular to the silicon wafer surface, with a spacing of 5mm±0.1mm; after the white light source 6 has been preheated for 30 minutes, the data processing unit 5 sends a synchronous trigger signal, the main spectrometer 4 collects the silicon wafer's reflectance spectral intensity R0, and the monitoring spectrometer 7 collects the initial light intensity B0 of the light source 6; then the silicon wafer reflectance fitting function f0 (N0,K0,THK0) is called, A0 is calculated according to A0=R0 / f0 (N0,K0,THK0), and K0 is obtained and stored. No repeated calibration is required during the actual measurement stage: the film 2 to be measured is placed in the same position on the stage, the data processing unit 5 sends a synchronous trigger signal at a set frequency, the main spectrometer 4 and the monitoring spectrometer 7 synchronously collect R and Bt, calculate Kt and the intensity of the incident light after calibration At=A0×Kt in real time, and finally obtain the thickness of the film 2 by fitting based on the principle of reflection spectroscopy.

[0045] Example 2: A method to improve the accuracy of thin film thickness measurement by reflectance spectroscopy Please see Figure 2 The method for improving the accuracy of thin film thickness measurement by reflectance spectroscopy disclosed in this embodiment is based on the system described in Embodiment 1. Through the process of "initial calibration - actual measurement - data processing", it solves the problems of low efficiency of repeated calibration and insufficient compensation for light source fluctuations in traditional methods.

[0046] The initial calibration step is crucial for establishing baseline parameters and only needs to be performed once. First, prepare a standard silicon wafer and clean its surface with a lint-free cloth dampened with isopropanol to remove contaminants. Place the cleaned silicon wafer on the measurement stage and secure it with positioning pins, ensuring the center of the silicon wafer is aligned with the center of the light spot on the multi-core fiber 9, with a deviation ≤0.1mm. Next, turn on the white light source 6, set the output power to 30mW, and preheat for 30 minutes to stabilize the light intensity. Simultaneously, adjust the integration time of the main and monitoring spectrometers 7 to 100ms to ensure the signal-to-noise ratio of the CCD chip is ≥50:1. Then, in the system software, click "Initial Calibration". The data processing unit 5 sends a synchronous trigger signal, the main spectrometer 4 collects the reflectance spectral intensity R0 of the silicon wafer in the 190-2300nm band, and the monitoring spectrometer 7 synchronously collects the light intensity B0 of the initial light source 6. The software automatically calls the silicon wafer reflectance fitting function f0(N0,K0,THK0) and calculates the initial incident light intensity A0 according to A0=R0 / f0(N0,K0,THK0). Then, the initial calibration coefficient K0 is obtained through K0=B0 / A0. A0, B0, K0, and other parameters such as calibration time and ambient temperature (25℃±2℃) are stored in the database.

[0047] The actual measurement steps enable continuous batch measurements without repeated calibration. First, the surface of the film to be measured 2 is cleaned using the same method as a standard silicon wafer. The silicon wafer on the stage is removed, and the film to be measured 2 is placed in the same position, ensuring alignment using positioning pins. The measurement chamber door is closed to avoid ambient light interference (ambient light influence ≤1%). The measurement batch (e.g., 50 pieces) and measurement frequency (e.g., 2Hz) are set in the system software. After clicking "Start Measurement," the stage automatically transports the film to be measured 2 to the measurement position. The data processing unit 5 sends a synchronous trigger signal at the set frequency. Each time a trigger occurs, the main spectrometer 4 acquires the reflected spectral intensity R, and the monitoring spectrometer 7 synchronously acquires the light intensity Bt of the current light source 6. After acquisition, the stage automatically transports the next film 2, achieving continuous batch measurements.

[0048] The core of the data processing and thickness calculation steps is dynamic calibration and real-time compensation. First, the acquired Bt is preprocessed using a 5-point moving average filter to remove high-frequency noise. Then, B0 and K0 from the database are retrieved, and the real-time calibration coefficient Kt is calculated using Kt = (Bt / B0) × K0. This coefficient reflects the intensity change of light source 6 at time t relative to the initial calibration time, achieving dynamic correlation. Next, the calibrated incident light intensity At is calculated using At = A0 × Kt. At is updated in real-time with Bt, compensating for instantaneous fluctuations and long-term attenuation of light source 6. Finally, based on the core model of the reflectance spectroscopy method, R / (A0×Kt)=f(N,K,THK) (R is the reflectance spectral intensity, f(N,K,THK) is the reflectance fitting function of the tested film 2, N and K are known optical constants, and THK is the thickness of film 2), the Levenberg-Marquardt fitting algorithm is used to solve for THK, and the fitting residual is ≤0.5%. The software stores data such as THK, measurement time, and Bt, and judges whether the tested film 2 is qualified according to the preset thickness range.

[0049] To ensure the time correlation of data acquisition, the synchronization of real-time acquisition of Bt and R acquired by the main spectrometer 4 is achieved in the following way: the GPIO interface of the data processing unit 5 outputs a 5V TTL synchronous trigger signal, which is simultaneously transmitted to the external trigger terminals of both spectrometers; after receiving the signal, the spectrometer starts acquisition with a delay of ≤1μs, and the acquisition duration is consistent with the integration time; after the acquisition is completed, the data is transmitted to the data processing unit 5 via USB 3.0 with a delay of <10ms, ensuring that R and Bt correspond one-to-one in time, and avoiding the impact of instantaneous fluctuations within the acquisition interval of the light source 6 on the calibration accuracy.

[0050] The above system and method were used to test and verify the thickness of a 100nm±5nm SiO2 standard thin film 2: the standard deviation of the thickness after 10 consecutive measurements was 0.32nm, which is about 16 times higher than the 5.1nm of the traditional method (without real-time calibration); the total time for batch measurement of 50 films 2 was 25 minutes, which is 60% more efficient than the 40 minutes of the traditional method; after 1000 hours of continuous use, the measurement deviation was only 0.2%, which is much lower than the 3.5% of the traditional method, fully demonstrating the advantages of the system and method in improving measurement accuracy, efficiency and long-term stability.

[0051] This article uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. It should be noted that those skilled in the art can make several improvements and modifications to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A system for improving the accuracy of thin film thickness measurement by reflectance spectroscopy, characterized in that, include: White light sources are used to emit light across a wide spectrum. The main spectrometer is used to measure the intensity of the reflected spectrum of the object being measured. A monitoring spectrometer is used to monitor the spectral intensity of a light source in real time. The optical fiber assembly includes a multi-core optical fiber and a single-core optical fiber. The single-core optical fiber is used to transmit part of the spectral signal of the light source to the monitoring spectrometer, and the multi-core optical fiber is used to transmit light from the light source to the object under test and to transmit the reflected light to the main spectrometer. The data processing unit is used to calibrate the incident light intensity in real time based on data from the main spectrometer and the monitoring spectrometer, and to calculate the film thickness.

2. The system according to claim 1, characterized in that, The multi-core optical fiber transmits light to the film being measured. The light reflected back from the film passes through one of the single cores of the multi-core optical fiber, transmitting the reflected spectral signal to the main spectrometer.

3. The system according to claim 1, characterized in that, The monitoring spectrometer has the same configuration and chip as the main spectrometer to ensure consistent signal acquisition.

4. The system according to claim 1, characterized in that, The data processing unit calculates the calibration coefficient Kt using the following formula: Kt = (Bt / B0) × K0; Where Bt is the light intensity measured by the monitoring spectrometer at time t, B0 is the reflection spectral intensity of the standard thickness silicon wafer, and K0 is the initial calibration coefficient.

5. The system according to claim 4, characterized in that, The initial calibration coefficient K0 is obtained by the following formula: K0 = B0 / A0; Where A0 is the incident light intensity obtained through standard silicon wafer calibration.

6. A method for improving the accuracy of thin film thickness measurement by reflectance spectroscopy, characterized in that, Includes the following steps: Initial calibration was performed using a standard silicon wafer to obtain the initial incident light intensity A0 and the monitoring light intensity B0. In actual measurements, the light intensity Bt measured by the monitoring spectrometer is acquired in real time. Calculate the real-time calibration coefficient Kt based on Bt and B0; The incident light intensity is calibrated in real time using Kt; The film thickness is calculated based on the calibrated incident light intensity.

7. The method according to claim 6, characterized in that, The formula for calculating the real-time calibration coefficient Kt is as follows: Kt = (Bt / B0) × K0.

8. The method according to claim 6 or 7, characterized in that, The film thickness is calculated based on the following model: R / (A0×Kt)=f(N,K,THK), Where R is the reflectance spectral intensity, f(N,K,THK) is the corresponding reflectance fitting function of the measured film, N and K are optical constants, and THK is the film thickness.

9. The method according to claim 6, characterized in that, The data acquisition time of the main spectrometer and the monitoring spectrometer is kept consistent.

10. The method according to claim 6, characterized in that, The real-time acquisition of the light intensity Bt measured by the monitoring spectrometer is synchronized with the reflectance spectrum signal acquired by the main spectrometer, specifically including: The data processing unit sends a synchronization trigger signal to the main spectrometer and the monitoring spectrometer; The main spectrometer and the monitoring spectrometer synchronously acquire a frame of reflectance spectrum signal and a frame of light source monitoring signal, respectively, based on the synchronization trigger signal, to obtain R and Bt at that moment.

Images