Method for detecting thickness of thin film through terahertz

Abstract

The invention provides a method for detecting the thickness of a thin film by terahertz. The method comprises the following steps: step 1, detecting a thick thin film: carrying out terahertz detection on the thick thin film to obtain a time domain signal; step 2, extracting a function model of the waveform of the pre-pulse signal: extracting a complete function model of the waveform of the pre-pulse signal from the time domain signal; 3, thin film detection: carrying out terahertz detection on the thin film to obtain a damaged time domain signal of the pre-pulse signal; step 4, fitting the waveform of the thin film; and step 5, calculating the thickness of the thin film: calculating the thickness of the thin film based on the flight time between the complete wave crest waveform of the pulse signal before restoration and the complete wave crest waveform of the pulse signal after detection. According to the invention, by processing the signal, the minimum detection thickness of the terahertz time-domain spectroscopy system is increased from more than 20 microns to several microns. The method is simple in principle, high in detection efficiency and accurate in result, and is expected to be applied to nondestructive high-precision detection of the thickness of the ultrathin material in actual engineering.

Description

Technical Field

[0001] The invention relates to the field of spectrum detection and signal processing, and in particular to a method for detecting film thickness by terahertz. Background Art

[0002] Nondestructive testing of the thickness of thin films, coatings or other ultra-thin objects is an important issue in engineering applications. Terahertz time-domain spectroscopy, which is developed based on terahertz waves, has high time resolution, as well as non-ionizing, non-destructive and non-contact characteristics, and is considered to be a promising tool for nondestructive evaluation of the thickness of thin coatings. Among them, terahertz time-domain spectroscopy can detect the terahertz time-domain signal of the sample, and the thickness of the sample can be quickly calculated using the time-of-flight method based on the time-domain signal. However, when detecting ultra-thin objects, the two pulse signals overlap because the distance between the front and rear pulse signals in the time-domain signal is too close, and the terahertz pulse signal itself has a certain pulse width, resulting in damage to the peak waveform of the front pulse signal, making it impossible to determine the flight time between the front and rear pulse signals. This makes the minimum thickness of the object that can be measured by terahertz time-domain spectroscopy combined with the time-of-flight method remain at more than 20 microns, while the thickness of some ultra-thin films, coatings and other objects is often only a dozen microns or even a few microns, which exceeds the detection limit of terahertz spectroscopy, limiting the application of terahertz detection technology in the field of ultra-thin films. Summary of the invention

[0003] In view of the shortcomings of the prior art, the present invention provides a method for detecting the thickness of thin films by terahertz. By processing the signal, the present invention increases the minimum detection thickness of the terahertz time-domain spectroscopy system from more than 20 microns to several microns. The method of the present invention has a simple principle, high detection efficiency and accurate results, and is expected to be applied to non-destructive and high-precision detection of the thickness of ultra-thin materials in practical engineering.

[0004] The present invention achieves the above technical objectives through the following technical means.

[0005] A method for detecting film thickness by terahertz, comprising the following steps:

[0006] Step 1: Detection of thick film: Perform terahertz detection on the thick film to obtain a time domain signal;

[0007] Step 2, extracting the function model of the pre-pulse signal waveform: extracting the complete function model of the pre-pulse signal waveform from the time domain signal;

[0008] Step 3: Thin film detection: perform terahertz detection on the thin film to obtain a time domain signal with a damaged pre-pulse signal;

[0009] Step 4, thin film waveform fitting: using the function model of the pre-pulse signal waveform extracted in step 2 to fit the damaged time domain signal of the pre-pulse signal in step 3, to restore the complete peak waveform of the pre-pulse signal;

[0010] Step 5, calculation of the thickness of the thin film: based on the flight time between the restored complete peak waveform of the front pulse signal and the detected complete peak waveform of the rear pulse signal, the thickness of the thin film is calculated.

[0011] In the above scheme, in step 1, terahertz time-domain spectroscopy is performed on the thick film to obtain its time-domain signal, and there will be two pulse signals in the time-domain signal; the first pulse signal appearing in the time domain is called the pre-pulse signal, and its waveform data is derived.

[0012] In the above scheme, in step 2, a model fitting program is used to extract model parameters from the exported pre-pulse signal waveform data.

[0013] In the above scheme, the model fitting program adopts a Gaussian function model, and fits the selected pre-pulse signal waveform through the Gaussian function model to extract its specific model parameters.

[0014] In the above scheme, in step three, terahertz time-domain spectroscopy is performed on the thin film, wherein the peak waveform of the front pulse signal of the thin film will be destroyed by the rear pulse signal, and the peak position of the front pulse signal cannot be identified, thereby obtaining a time domain signal with a damaged front pulse signal.

[0015] In the above scheme, the flight time in step 5 refers to the time interval between the only post-pulse signal peak in the thin film time domain signal and the pre-pulse signal peak fitted by the thin film. The specific formula for calculating the thickness of the thin film by the flight time method is: d is the thickness of the thin film, Δt is the time interval, n is the refractive index of the thin film, and c is the speed of light.

[0016] In the above scheme, the thick film and the thin film are made of the same material.

[0017] In the above scheme, the thickness of the thick film is greater than 30 μm.

[0018] In the above scheme, the thickness of the thin film is 4 μm to 25 μm.

[0019] Beneficial effects of the present invention:

[0020] 1) The present invention adopts a model fitting method to extract a specific Gaussian function model of the complete pre-pulse signal peak waveform from the time domain signal of the thick film, and applies it to the fitting of the incomplete pre-pulse signal peak waveform of the thin film, thereby obtaining the complete pre-pulse signal peak waveform of the thin film, and then determining the peak position.

[0021] 2) The method proposed in the present invention can greatly improve the detection level of the terahertz time-domain spectroscopy system, increasing its minimum detection thickness from more than 20 microns to several microns, thereby realizing the thickness detection of thin films.

[0022] 3) The signal processing method proposed in the present invention is implemented through programming, and the computer is used to complete the calculation of a large amount of data involved in the fitting process. It is simple, efficient and accurate in calculation, and can be applied to non-destructive and high-precision detection of the thickness of ultra-thin materials in actual engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 Schematic diagram of the time domain signal waveform measured for thick film;

[0024] Figure 2 A schematic diagram of extracting a specific function model of a thick film based on its time domain signal;

[0025] Figure 3 Schematic diagram of the time domain signal waveform measured for thin films;

[0026] Figure 4 This is a schematic diagram of a complete pre-pulse signal waveform obtained after model fitting from an incomplete pre-pulse signal waveform in the time domain signal of a thin film. DETAILED DESCRIPTION

[0027] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0028] Specifically, as shown in the accompanying drawings, a method for detecting film thickness by using terahertz time-domain spectroscopy with high efficiency and high precision according to an embodiment of the present invention includes:

[0029] Step 1: Thick film detection: Figure 1 As shown, the thick film is detected by terahertz to obtain its time domain signal;

[0030] Figure 1 In the case of thicker films, the signals detected by the terahertz time-domain spectroscopy system will have two pulse signals, the front pulse signal has a smaller overall amplitude than the rear pulse signal. The amplitude of a single complete pulse signal will rise first, then begin to decline after reaching the peak, and then rise again after falling to the lowest point below the horizontal line (the initial amplitude line without signal), and finally return to the horizontal line. Figure 1 The waveforms of the pulse signals before and after are consistent with the above-mentioned waveform changes, which means that the waveforms of the two are relatively complete and the positions of their respective peaks are identifiable.

[0031] Step 2: Extraction of pre-pulse model: Figure 2 As shown, the specific parameters of the Gaussian function model of the pre-pulse signal waveform in the time domain signal of the thick film are extracted;

[0032] Figure 2 In the method, the waveform data of the peak part of the pre-pulse signal in the time domain signal is selected, and the specific model parameters that best fit the Gaussian function model can be extracted through the fitting program, that is, the peak waveform of the pre-pulse signal is described by using a specific Gaussian function.

[0033] Step 3: Thin film testing: Figure 3 As shown, the time domain signal of the thin film is measured, in which the front pulse signal is affected by the rear pulse signal, the peak is partially damaged, resulting in an incomplete front pulse signal and the peak position cannot be identified.

[0034] Figure 3 It can be clearly seen from the measured time domain signal waveform of the thin film that the front pulse signal is interrupted by the appearance of the rear pulse signal in the initial amplitude rising stage, resulting in only part of the front pulse signal in the time domain signal of the thinner film. The front pulse signal is incomplete, especially the peak part disappears and the peak position cannot be identified.

[0035] Step 4: Fitting of thin film waveform: Figure 4 As shown, a model is fitted to the part of the time domain signal of the thinner film that belongs to the pre-pulse signal to obtain a complete peak waveform of the pre-pulse signal, and to determine the position of the peak and the flight time.

[0036] Figure 4 In the figure, it can be seen from the waveform of the time domain signal that the pre-pulse signal is incomplete, and only a part of the signal at the initial rise exists. The waveform data of this part of the signal is selected and imported into the fitting program, and then the Gaussian model function of the pre-pulse signal peak waveform extracted from the thicker same film time domain signal is input into the fitting program to fit the complete peak waveform data and waveform of the pre-pulse signal in the time domain signal of the thinner film, determine the peak position of the pre-pulse signal and calculate the flight time between the front and rear pulse signals.

[0037] Step 5, calculation of thin film thickness: Based on the flight time, the thickness of the thinner film can be quickly calculated through the classic flight time method, thereby realizing the thickness detection of the thinner film.

[0038] Example 1

[0039] First, terahertz time-domain spectroscopy was performed on the 108-micron PI film to obtain its time-domain signal.

[0040] Then, the Gaussian function model and the model fitting program written by the particle swarm algorithm are used to extract the Gaussian model parameters from the waveform data of the peak part of the front pulse signal in the time domain signal of the 108-micron PI film, and the specific function based on the Gaussian function model is obtained.

[0041] Then, the 23-micron PI film was subjected to terahertz time-domain spectroscopy detection to obtain its time-domain signal.

[0042] Then, the part of the signal waveform data belonging to the front pulse signal in the time domain signal waveform diagram measured by the 23-micron PI film is imported into the model fitting program, and the specific function of the front pulse signal peak waveform based on the Gaussian function model of the 108-micron PI film is filled into the model fitting program. By running the model fitting program, the complete peak waveform of the front pulse signal in the time domain signal measured by the 23-micron PI film can be obtained and the position of the peak can be determined, and then the flight time between the front and rear pulse signals is obtained to be 0.334ps. The fitting thickness of the 23-micron PI film is calculated to be 25.7 microns using the flight time method.

[0043] Example 2

[0044] First, terahertz time-domain spectroscopy was performed on the 108-micron PI film to obtain its time-domain signal.

[0045] Then, the Gaussian function model and the model fitting program written by the particle swarm algorithm are used to extract the Gaussian model parameters from the waveform data of the peak part of the front pulse signal in the time domain signal of the 108-micron PI film, and the specific function based on the Gaussian function model is obtained.

[0046] Then, the 14-micron PI film was subjected to terahertz time-domain spectroscopy detection to obtain its time-domain signal.

[0047] Then, the part of the signal waveform data belonging to the front pulse signal in the time domain signal waveform diagram measured by the 14-micron PI film is imported into the model fitting program, and the specific function of the front pulse signal peak waveform based on the Gaussian function model of the 108-micron PI film is filled into the model fitting program. By running the model fitting program, the complete peak waveform of the front pulse signal in the time domain signal measured by the 14-micron PI film can be obtained and the position of the peak can be determined, and then the flight time between the front and rear pulse signals is obtained to be 0.198ps. The fitting thickness of the 14-micron PI film is calculated to be 15.2 microns using the flight time method.

[0048] Example 3

[0049] First, terahertz time-domain spectroscopy was performed on the 108-micron PI film to obtain its time-domain signal.

[0050] Then, the Gaussian function model and the model fitting program written by the particle swarm algorithm are used to extract the Gaussian model parameters from the waveform data of the peak part of the front pulse signal in the time domain signal of the 108-micron PI film, and the specific function based on the Gaussian function model is obtained.

[0051] Then, the 9-micron PI film was subjected to terahertz time-domain spectroscopy detection to obtain its time-domain signal.

[0052] Then, the part of the signal waveform data belonging to the front pulse signal in the time domain signal waveform diagram measured by the 9-micron PI film is imported into the model fitting program, and the specific function of the front pulse signal peak waveform based on the Gaussian function model of the 108-micron PI film is filled into the model fitting program. By running the model fitting program, the complete peak waveform of the front pulse signal in the time domain signal measured by the 9-micron PI film can be obtained and the position of the peak can be determined, and then the flight time between the front and rear pulse signals is obtained to be 0.13ps. The fitting thickness of the 9-micron PI film is calculated to be 10 microns using the flight time method.

[0053] Example 4

[0054] First, terahertz time-domain spectroscopy was performed on the 108-micron PI film to obtain its time-domain signal.

[0055] Then, the Gaussian function model and the model fitting program written by the particle swarm algorithm are used to extract the Gaussian model parameters from the waveform data of the peak part of the front pulse signal in the time domain signal of the 108-micron PI film, and the specific function based on the Gaussian function model is obtained.

[0056] Then, the 7-micron PI film was subjected to terahertz time-domain spectroscopy detection to obtain its time-domain signal.

[0057] Then, the part of the signal waveform data belonging to the front pulse signal in the time domain signal waveform diagram measured by the 7-micron PI film is imported into the model fitting program, and the specific function of the front pulse signal peak waveform based on the Gaussian function model of the 108-micron PI film is filled into the model fitting program. By running the model fitting program, the complete peak waveform of the front pulse signal in the time domain signal measured by the 7-micron PI film can be obtained and the position of the peak can be determined, and then the flight time between the front and rear pulse signals is obtained to be 0.097ps. The fitting thickness of the 7-micron PI film is calculated to be 7.5 microns using the flight time method.

[0058] In combination with Examples 1-4, the film thickness obtained by the method of the present invention is close to the actual thickness. According to the above data, the relative error of the fitting result of the film thickness is maintained at 7% to 12%, indicating that the actual thickness of the film can be measured more accurately by the method of the present invention, and the detection limit of the thickness measured by the terahertz time-of-flight method is reduced to (25μm to 7μm) by the method of the present invention.

[0059] Example 5

[0060] TGO film

[0061] First, terahertz time domain signal detection is performed on a 204-micron thick aluminum oxide thin block with the same material composition as the TGO film to obtain its time domain signal.

[0062] Then, the time domain signal of the aluminum oxide thin block is processed by a model fitting program written using a Gaussian function model and a particle swarm algorithm, and the waveform data of the front pulse signal in the signal is subjected to Gaussian model parameter extraction to obtain a specific function based on the Gaussian function model.

[0063] Then, the 4.8-micron TGO film sample was subjected to terahertz time-domain spectroscopy detection to obtain its time-domain signal.

[0064] Then, the part of the signal waveform data belonging to the front pulse signal in the time domain signal waveform diagram measured by the 4.8 micron TGO coating is imported into the model fitting program, and the specific function of the front pulse signal peak waveform based on the Gaussian function model of the 204 micron alumina block is filled into the model fitting program. By running the model fitting program, the complete peak waveform of the front pulse signal in the time domain signal measured by the 4.8 micron TGO coating can be obtained and the position of the peak can be determined, and then the flight time between the front and rear pulse signals is obtained to be 0.0759ps. The fitting thickness of the 4.8 micron TGO coating is calculated to be 3.93 microns using the flight time method.

[0065] Example 6

[0066] TGO film

[0067] First, terahertz time domain signal detection is performed on a 204-micron thick aluminum oxide thin block with the same material composition as the TGO film to obtain its time domain signal.

[0068] Then, the time domain signal of the aluminum oxide thin block is processed by a model fitting program written using a Gaussian function model and a particle swarm algorithm, and the waveform data of the front pulse signal in the signal is subjected to Gaussian model parameter extraction to obtain a specific function based on the Gaussian function model.

[0069] Then, the 4-micron TGO film sample was subjected to terahertz time-domain spectroscopy detection to obtain its time-domain signal.

[0070] Then, the part of the signal waveform data belonging to the front pulse signal in the time domain signal waveform diagram measured by the 4-micron TGO coating is imported into the model fitting program, and the specific function of the front pulse signal peak waveform based on the Gaussian function model of the 204-micron alumina block is filled into the model fitting program. By running the model fitting program, the complete peak waveform of the front pulse signal in the time domain signal measured by the 4-micron TGO coating can be obtained and the position of the peak can be determined, and then the flight time between the front and rear pulse signals is obtained to be 0.0635ps. The fitting thickness of the 4-micron TGO coating is calculated to be 3.28 microns using the time-of-flight method.

[0071] Combining Examples 5 and 6, the TGO coating thickness is relatively close to its actual thickness, with an actual error of about 0.8 μm, and the minimum detected thickness is 3 to 4 μm.

[0072] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0073] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention without departing from the principles and intent of the present invention.

Claims

1. A method for detecting film thickness using terahertz, characterized in that: The specific steps include: Step 1: Detection of thick film: Perform terahertz detection on the thick film to obtain a time domain signal; Step 2, extracting the function model of the pre-pulse signal waveform: extracting the complete function model of the pre-pulse signal waveform from the time domain signal; Step 3: Thin film detection: perform terahertz detection on the thin film to obtain a time domain signal with a damaged pre-pulse signal; Step 4, thin film waveform fitting: using the function model of the pre-pulse signal waveform extracted in step 2 to fit the damaged time domain signal of the pre-pulse signal in step 3, to restore the complete peak waveform of the pre-pulse signal; Step 5, calculation of the thickness of the thin film: based on the flight time between the restored complete peak waveform of the front pulse signal and the detected complete peak waveform of the rear pulse signal, the thickness of the thin film is calculated.

2. The method for detecting film thickness by terahertz according to claim 1, characterized in that: In the step 1, the thick film is subjected to terahertz time-domain spectroscopy detection to obtain its time-domain signal, in which there are two pulse signals; the first pulse signal appearing in the time domain is called the pre-pulse signal, and its waveform data is derived.

3. The method for detecting film thickness by terahertz according to claim 1, characterized in that: In the step 2, a model fitting program is used to extract model parameters from the exported pre-pulse signal waveform data.

4. The method for detecting film thickness by terahertz according to claim 3, characterized in that: The model fitting program adopts a Gaussian function model, fits the selected pre-pulse signal waveform through the Gaussian function model, and extracts its specific model parameters.

5. The method for detecting film thickness by terahertz according to claim 1, characterized in that: In the step three, terahertz time-domain spectroscopy is performed on the thin film, wherein the peak waveform of the front pulse signal of the thin film will be destroyed by the rear pulse signal, and the peak position of the front pulse signal cannot be identified, thereby obtaining a time domain signal with a damaged front pulse signal.

6. The method for detecting film thickness by terahertz according to claim 1, characterized in that: The flight time in step 5 refers to the time interval Δt between the only post-pulse signal peak in the thin film time domain signal and the pre-pulse signal peak fitted by the thin film. The specific formula for calculating the thickness of the thin film by the flight time method is: d is the thickness of the thin film, Δt is the time interval, n is the refractive index of the thin film, and c is the speed of light.

7. The method for detecting film thickness by terahertz according to claim 1, characterized in that: Thick film and thin film are made of the same material.

8. The method for detecting film thickness by terahertz according to claim 1, characterized in that: Thick film thickness is greater than 30μm.

9. The method for detecting film thickness by terahertz according to claim 1, characterized in that: The thickness of the thin film is 4μm~25um.

Images