Method for fitting ellipsometry measurement data of continuous gradient component film

Abstract

The invention provides a method for fitting ellipsometry measurement data of a continuous gradient component film, which is characterized by comprising the following specific steps of: S1, selecting a plurality of measurement points at equal intervals for ellipsometry test; s2, carrying out advanced ellipsometry data analysis on a test point close to the central position in all the measurement points to obtain an inversion result; s3, determining the quality of the inversion result of the test point close to the central position by using an evaluation function; s4, performing ellipsometry data analysis on the test points on the two sides of the test point close to the central position according to the optimal ellipsometry fitting result of the test point close to the central position obtained by the evaluation function; and S5, carrying out ellipsometry data analysis and evaluation on adjacent unanalyzed test points until the optimal fitting parameters including the film thickness and the dielectric functions of all ellipsometry measurement data points are obtained. According to the method for fitting the continuous gradient component film, the workload of ellipsometry in the aspect of data fitting is reduced, and the dielectric functions are closely associated but have respective characteristics.

Description

Technical Field

[0001] The invention relates to the technical field of thin film ellipsometric measurement, and in particular to a method for fitting continuous gradient component thin film ellipsometric measurement data. Background Art

[0002] Continuous gradient component thin film refers to a thin film sample that uses high-throughput preparation technology to cover a wide range of components on the same substrate at one time. Its advantages are that the growth and measurement conditions are almost completely consistent. On the other hand, since the same continuous gradient component film contains a wide range of components, it can greatly shorten the research and development cycle of new materials. Based on this feature of continuous gradient component thin films, the characterization method must have extremely high resolution. Ellipsometer spectrometer (ellipsometer) is very suitable for optical measurement of continuous gradient component thin films because the light spot can be focused and it is a characterization method that integrates multiple advantages such as high precision, high sensitivity, and non-destructiveness. In ellipsometry (ellipsometry), the Fresnel reflection coefficients of p-light and s-light in the incident linear polarized light are different, resulting in changes in the amplitude and phase of the two beams of reflected light. Therefore, the test data is the amplitude ratio of the ellipsometric angle p-light and s-light tanΨ and the phase difference cosΔ between p-light and s-light. If you want to obtain the dielectric function of interest ε r and ε i , the dispersion equation needs to be used for inversion fitting to obtain it.

[0003] However, the current mature fitting methods are only for single-component films. If the method of fitting single-component films is used to process the ellipsometric data of continuous gradient component films, there are the following problems: (1) Due to the dramatic increase in the amount of ellipsometric measurement data of continuous gradient component films, the method of constructing a dispersion model alone and fitting the ellipsometric data of the measurement points point by point will greatly increase the workload of obtaining the dielectric function; (2) The fitting method of single-component films is independent (this is because for single-component samples with the same composition, even if the preparation conditions between samples are exactly the same, there are still inevitable differences in parameters such as composition and film thickness). The initial value of the established dispersion model has nothing to do with the film thickness, which causes the dielectric functions between data points to be independent of each other, which will bring considerable errors to the subsequent physical analysis, such as the physical analysis of the plasma oscillation frequency and scattering rate at each point, the calculation of photoconductivity (resistivity) and other physical problems; (3) There are only a few reports on the ellipsometric fitting data of continuous gradient component films, which do not fully highlight the application of ellipsometric measurement in continuous gradient component films, limiting the research on continuous gradient component films in optics.

[0004] For example, the invention patent with application number 202311010723.5 provides a method, device, and equipment for determining the intrinsic bandgap width of a thin film dielectric material, wherein the method includes: performing ellipsometry measurement on a single crystal or polycrystalline thin film dielectric material to be measured to obtain ellipsometry measurement parameters; generating ellipsometry simulation parameters based on a pre-built measurement model; iterating and fitting the ellipsometry measurement parameters and the ellipsometry simulation parameters to generate a spectral absorption dispersion curve; performing Gaussian function peak fitting processing based on the spectral absorption dispersion curve to obtain peak fitting Gaussian function parameters and reference dispersion curves; extracting the target fitting Gaussian function parameters and target reference dispersion curves corresponding to the characteristic absorption peaks of the thin film dielectric material and performing normalization and linear extrapolation to determine the intrinsic bandgap width of the thin film dielectric material. The intrinsic bandgap width of the single crystal or polycrystalline thin film dielectric material to be measured is determined by normalizing and linearly extrapolating the target fitting Gaussian function parameters and the target reference dispersion curve. However, this method is relatively complex and requires a large amount of continuous gradient data. Summary of the invention

[0005] In view of the technical problems that the existing ellipsometric data fitting methods of continuous gradient component films have large errors and large calculation amount, the present invention proposes a method for fitting the ellipsometric measurement data of continuous gradient component films, which reduces the workload of ellipsometric measurement in data fitting and avoids a large amount of error accumulation in the fitting process.

[0006] In order to achieve the above object, the technical solution of the present invention is implemented as follows: a method for fitting continuous gradient component thin film ellipsometry measurement data, the specific steps are:

[0007] S1: Select multiple measurement points at equal intervals for ellipsometric testing, and use an ellipsometer to obtain the ellipsometric angles tanΨ and cosΔ of all measurement points;

[0008] S2: Using the joint dispersion model including the Drude equation and the Lorentz oscillator equation, the ellipsometry data of the test points close to the center of all the measurement points are analyzed in advance to obtain the inversion results;

[0009] S3: Fitting the joint dispersion model, and using the evaluation function to determine the quality of the inversion results of the test points close to the center position, to obtain the parameters of the test points close to the center position;

[0010] S4: performing ellipsometric data analysis on the test points on both sides of the test point close to the center position according to the evaluation function to obtain the best ellipsometric fitting result of the test point close to the center position;

[0011] S5: After completing the ellipsometric data analysis of the test points on both sides, perform ellipsometric data analysis and evaluation on the test points adjacent to the test points on both sides that have not been analyzed, until the best fitting parameters including the film thickness and the dielectric function of all the ellipsometric measurement data points are obtained.

[0012] The joint dispersion model including the Drude equation and the Lorentz oscillator equation is:

[0013]

[0014]

[0015] Among them, ε r is the real part of the dielectric function, ε i is the imaginary part of the dielectric function, λ is the wavelength of the incident light, P is the electric polarization intensity in the Drude dispersion equation, D is the inverse of the plasma wavelength in the Drude dispersion equation, E is the mean free path of electron motion in the Drude dispersion equation, A is the peak intensity of the Lorentz oscillator, L0 is the central wavelength of the Lorentz, and γ is the broadening of the Lorentz.

[0016] The method for obtaining the inversion result in step S2 is: performing ellipsometric inversion using simulated annealing-ant colony algorithm, artificial ant colony algorithm or point-to-point algorithm to obtain the inversion result.

[0017] The evaluation function is:

[0018]

[0019] Where n is the number of test points within the test range, m is the number of parameters involved in the fitting process, Ψ is the amplitude ratio and phase difference of the p-light before and after reflection, Δ is the amplitude ratio and phase difference of the s-light before and after reflection, and the subscripts cal and exp represent the calculated value and experimental value of the ellipsometric angle, respectively.

[0020] The method for fitting the joint dispersion model in step S3 is: fitting the fitting parameters of the joint dispersion model using a simulated annealing algorithm.

[0021] The fitting parameters are the electric polarization intensity P, the inverse of the plasma wavelength D, the mean free path E of the electron movement, the peak intensity A of the Lorentz oscillator, the central wavelength L0 of the Lorentz and the broadening γ of the Lorentz.

[0022] The method of using the evaluation function to evaluate the quality of the inversion result in step S3 is:

[0023] Input the inversion result into the evaluation function. If the output of the evaluation function is in the order of magnitude of 10 -4 If the magnitude of the evaluation function output result is greater than 10 -4 If the inversion result is not satisfactory, it is necessary to repeat step S2 to obtain the inversion result again.

[0024] The test point parameters in step S3 are the fitting parameters, film thickness value and dielectric function of the test point close to the center position.

[0025] The method for performing ellipsometric data analysis on the test points on both sides of the test point close to the center position in step S4 is:

[0026] S41: performing ellipsometry data analysis using test points on both sides of a joint dispersion model including the Drude equation and the Lorentz oscillator equation, and setting the joint dispersion model including the Drude equation and the Lorentz oscillator equation and a film thickness value;

[0027] S42: Obtain the inversion results of the two side points through ellipsometry data analysis, input the inversion results of the two side points into the evaluation function for evaluation, and if the output result of the evaluation function is within the order of magnitude of 10 -4 If the magnitude of the evaluation function output result is greater than 10 -4 If the inversion result is not satisfactory, the fitting parameters need to be adjusted and the ellipsometric data analysis needs to be repeated.

[0028] The method for setting the joint dispersion model including the Drude equation and the Lorentz oscillator equation and the film thickness value described in step S41 is: using the best ellipsometry fitting result as the initial value to set the electric polarization intensity P in the Drude dispersion equation, the inverse of the plasma wavelength D in the Drude dispersion equation, the mean free path E of the electron motion in the Drude dispersion equation, the peak intensity A of the Lorentz oscillator, the central wavelength L0 of the Lorentz, the broadening γ of the Lorentz and the film thickness value in the joint dispersion model including the Drude equation and the Lorentz oscillator equation.

[0029] The advantages of the present invention over the prior art are: a new analysis method is proposed for the ellipsometric data fitting of continuous gradient component films, which not only greatly reduces the workload of ellipsometric measurement in data fitting, but also ensures that the dielectric functions are closely related while retaining their respective characteristics. More importantly, a large amount of error accumulation is not caused during the fitting process, and the breadth and depth of optical research on continuous gradient component films is greatly expanded. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0031] Figure 1The figure is a flow chart of the method of the present invention.

[0032] Figure 2 Schematic diagram of ellipsometric testing of continuous gradient composition films.

[0033] Figure 3 Ellipsometric test curve and fitting curve for all measurement points.

[0034] Figure 4 It is the evaluation function diagram of each measured data point of the continuous gradient component film.

[0035] Figure 5 It is the dielectric function diagram of each measured data point of the continuous gradient component film. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] Example 1

[0038] like Figure 1 As shown, a method for fitting continuous gradient component thin film ellipsometry measurement data, the specific steps are:

[0039] S1: Select multiple measurement points at equal intervals for ellipsometric testing, and use an ellipsometer to obtain the ellipsometric angles tanΨ and cosΔ of all measurement points.

[0040] When selecting test points, it is necessary to ensure that the detected component has a continuous gradient at each measuring point, and there is no specific restriction on the positional relationship between the test points.

[0041] like Figure 2 This is a schematic diagram of the ellipsometric test of a continuous gradient component film. The yellow solid circle in the figure is the location of the micro spot. Since the composition of a certain element in the sample changes continuously from 1 to 16 (the number of test points is only for reference), the measurement order is 1-16.

[0042] S2: Use the joint dispersion model including the Drude equation and the Lorentz oscillator equation to perform preliminary ellipsometric data analysis on the test points close to the center position of all measurement points.

[0043] Specifically, the joint dispersion model including the Drude equation and the Lorentz oscillator equation includes one Drude equation and two Lorentz oscillator equations:

[0044]

[0045]

[0046] Among them, ε r and ε i represent the real and imaginary parts of the dielectric function respectively; λ is the wavelength of incident light, P is the electric polarization intensity in the Drude dispersion equation, D is the inverse of the plasma wavelength in the Drude dispersion equation, E is the mean free path of electron motion in the Drude dispersion equation, P, D and E are all fitting parameters, A is the peak intensity of the Lorentz oscillator, L0 is the central wavelength of the Lorentz, γ is the broadening of the Lorentz, A, L0 and γ are all fitting parameters.

[0047] P, D, E, A, L0 and γ in the joint dispersion model including the Drude equation and the Lorentz oscillator equation are set, and the initial value of the film thickness is set, and ellipsometry inversion is performed to obtain the inversion result. The inversion calculation algorithm includes but is not limited to simulated annealing-ant colony algorithm, artificial ant colony algorithm, point-to-point algorithm, etc.

[0048] Assume that in step S1 according to Figure 2 If 16 test points are selected, then in step S2 09 test points may be selected for ellipsometric inversion.

[0049] The Drude and Lorentz dispersion models use equations to describe the variation of the dielectric function of the material to be tested with wavelength. Drude is used to describe the influence of low-frequency carriers on the dielectric function, and Lorentz is used to describe the characteristics of inter-band transitions of electrons.

[0050] S3: Use the evaluation function to determine the quality of the dispersion model inversion results involving the fitting parameters and film thickness values ​​of the test points close to the center.

[0051] Specifically, the fitting parameters of the joint dispersion model including the Drude equation and the Lorentz oscillator equation are fitted using a simulated annealing algorithm, and the inversion results are evaluated using an evaluation function to obtain the best fitting parameters, film thickness values, and dielectric functions of the test points close to the center.

[0052] Specifically, the evaluation function is:

[0053]

[0054] Where n is the number of test points within the test range, m is the number of parameters involved in the fitting process, Ψ is the amplitude ratio and phase difference of the p-light before and after reflection, Δ is the amplitude ratio and phase difference of the s-light before and after reflection, and the subscripts cal and exp represent the calculated value and experimental value of the ellipsometric angle, respectively. This is tanΨ obtained in step S1.

[0055] If the output of the evaluation function is in the order of 10 -4 If the magnitude of the evaluation function output result is greater than 10, it is the best ellipsometric fitting result and can proceed to the next step. -4 , the inversion result is not ideal, and the fitting parameters need to be adjusted, and step S2 is repeated to obtain the inversion result again.

[0056] S4: performing ellipsometric data analysis on the test points on both sides of the test point close to the center position according to the evaluation function to obtain the best ellipsometric fitting result of the test point close to the center position.

[0057] Specifically, the test points on both sides of the joint dispersion model including the Drude equation and the Lorentz oscillator equation are used for ellipsometric data analysis, and the P, D, E, A, L0, γ and film thickness values ​​in the joint dispersion model including the Drude equation and the Lorentz oscillator equation are set with the best ellipsometric fitting result as the initial value. The inversion results of the points on both sides are obtained by ellipsometric data analysis, and the inversion results of the points on both sides are input into the evaluation function for evaluation. If the output result of the evaluation function is in the order of magnitude of 10 -4 If the magnitude of the evaluation function output result is greater than 10, it is the best ellipsometric fitting result and can proceed to the next step. -4 , the inversion result is not ideal and the fitting parameters need to be adjusted and the ellipsometric data analysis needs to be repeated.

[0058] If in step S1 according to Figure 2 16 test points are selected, and test point 09 can be selected for ellipsometric inversion in step S2. After obtaining the best ellipsometric fitting result of test point 09, ellipsometric data analysis is performed on test points 08 and 10. During the ellipsometric data analysis, the initial fitting parameters and film thickness values ​​in the joint dispersion model including the Drude equation and the Lorentz oscillator equation are consistent with the best ellipsometric fitting result of test point 09.

[0059] S5: After completing the ellipsometric data analysis of the test points on both sides, perform ellipsometric data analysis and evaluation on the unanalyzed test points adjacent to the test points on both sides until the best fitting parameters (including film thickness) and dielectric function of all ellipsometric measurement data points are obtained.

[0060] Ellipsometry measurement and fitting of single-component thin films is a very mature technology at present. Since the composition of single-component samples is uniform, it is only necessary to select a certain point on the sample surface for ellipsometric testing. When fitting the ellipsometric angle tanΨ and cosΔ using the inversion method, select an appropriate dispersion equation, input the initial value of the parameter to be fitted in the dispersion equation, and continuously iterate to obtain the optimal solution, and then obtain the dielectric function ε rand ε i . On the surface of the continuous gradient component thin film sample, the composition of a certain chemical element changes continuously from one end of the substrate to the other end, so the physical properties at different points are different. If the currently mature ellipsometric fitting method for single-component thin film samples is used, each point needs to be considered isolated and unrelated to each other. This has certain problems, that is, when the measured data points are considered to be unrelated, the selection of the initial value of the dispersion model is also unrelated, which results in certain differences between the dielectric functions finally obtained. Therefore, for this fitting method, the workload is large and the results lose accuracy.

[0061] In fact, the composition of the continuous gradient component film changes gradually in a certain direction, and its optical properties (including other physical properties) also change continuously and gradually with the composition, while the optical properties of non-adjacent data points have sudden changes. Therefore, the fitting method of the present application is to select a certain point in the test point for preliminary fitting, and the adjacent data points are fitted using the fitting result of this point as the initial reference, including selecting the same dispersion model and initial values ​​of fitting parameters. It is explained here that the preliminary fitting point of this work is selected in the middle position of the ellipsometric measurement point, in order to ensure that the error caused by the step of gradually fitting from the middle to both sides is not too large; if the preliminary fitting point is selected at the first or last point, the error in the fitting process will gradually accumulate, and eventually even cause the dielectric function to deviate seriously from its true value. The method of fitting continuous gradient component films proposed in the present application not only greatly reduces the workload of ellipsometric measurement in data fitting, but also the dielectric functions are closely related without losing their respective characteristics.

[0062] Example 2

[0063] A method for fitting ellipsometric data of a continuous gradient component film. In this embodiment, a continuous gradient component film LaCe x Cu 1-x Taking O4 (LCCO, 0.1≤x≤0.19) as an example, a method for fitting the ellipsometry data of a continuous gradient composition thin film is described, comprising the following steps:

[0064] S1: Select measurement points with equal intervals for ellipsometric test, for example, 16 data points in total. Figure 2 As shown in Figure 2, the ellipsometer is used to obtain the ellipsometric angles tanΨ and cosΔ of all measurement points, as shown in Figure 2 Figure 3 As shown by the hollow circles in;

[0065] Specifically, the ellipsometer GES-5 produced by SOPRA, France, was used to measure the ellipsometric process of the continuous gradient composition thin film LCCO. The measurement temperature was room temperature, the incident angle was fixed at 70°, and the measured wavelength range was 300-800nm.

[0066] S2: Select a point in the middle of all the measurement points for preliminary ellipsometric data analysis, such as point 09. The purpose of this is to avoid the error accumulation caused by the preliminary fitting from the first point. Construct a joint dispersion model including the Drude equation and the Lorentz oscillator equation, including 1 Drude equation and two Lorentz oscillator equations, and set the initial values ​​of the dispersion model fitting parameters and the film thickness to perform ellipsometric inversion;

[0067]

[0068]

[0069] Where λ is the wavelength of the incident light; P, D and E represent the electric polarization intensity, the inverse of the plasma wavelength and the mean free path of electron motion in the Drude dispersion equation, respectively, and are all fitting parameters; A, L0 and γ represent the peak intensity, central wavelength and broadening of the Lorentz oscillator, respectively, and are all fitting parameters.

[0070] S3: Use the evaluation function to determine the quality of the dispersion model inversion results involving the fitting parameters (including film thickness) of the 09th test point, such as Figure 4 As shown, the evaluation function is in the order of 10 -4 The magnitude is an excellent fitting result. The fitting curve is as follows Figure 3 As shown by the solid line in .

[0071] The film thickness in the fitting result is 86nm, while the actual thickness measured by scanning electron microscopy is 85nm, with only a 1% error.

[0072] Among them, the algorithms for thin film optical inversion calculation include regression algorithm, ant colony algorithm, simulated annealing algorithm and so on.

[0073] When the evaluation function is the smallest, it is considered that the corresponding best fitting result is obtained. The evaluation function is as follows:

[0074]

[0075] n represents the number of test points within the test range, m represents the number of all parameters involved in the fitting process, Ψ and Δ represent the amplitude ratio and phase difference of p-light and s-light before and after reflection, respectively, and the subscripts cal and exp represent the calculated value and experimental value of the ellipsometric angle, respectively;

[0076] The best dispersion model fitting parameters are shown in Table 1;

[0077] Table 1 Taking the 09th test point of the continuous gradient component film as an example, the best fitting parameters of the joint dispersion model are listed

[0078]

[0079]

[0080] For the 08th test point, all fitting-related information is no longer set separately, but the result of the 09th test point is used as the reference value, including the dispersion model, the fitting parameters in the dispersion model, and the film thickness information; similarly, the ellipsometry data inversion of the 10th test point is also based on the result of the 09th test point;

[0081] The evaluation function magnitude of the two points is also 10 -4 , the fitting curve is as follows Figure 3 As shown by the solid line in .

[0082] S4: The initial values ​​of the ellipsometric parameters at the 7th and 11th measuring points are respectively based on the results of the 8th and 10th measuring points. The evaluation function and fitting curve of the obtained results are as follows: Figure 3 Solid line and Figure 4 shown. Figure 3 In the figure, the hollow circles are the measured data, the solid line is the fitting curve after the steps of constructing the dispersion model and inversion iteration, and the black arrow represents the trend of the Ce content in LCCO gradually increasing from 0.1 to 0.19.

[0083] S5: Using this method, and so on, the best fitting curve of all measurement points is finally obtained ( Figure 3 Solid line) and the variation of dielectric function with wavelength ( Figure 5 ).

[0084] The final obtained curve of the change of dielectric function with wavelength has similar behavioral trends between the dielectric functions of each measurement point, but also contains their own curve characteristics, which provides accurate data guarantee for the subsequent in-depth analysis of the physical mechanism.

[0085] The other structures and principles are the same as those in Example 1.

[0086] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for fitting continuous gradient composition thin film ellipsometry data, characterized in that: The specific steps are: S1: Select multiple measuring points at equal intervals on the continuous gradient component film for ellipsometry testing to obtain the ellipsometric angles tanΨ and cosΔ of all measuring points; S2: Using the joint dispersion model including the Drude equation and the Lorentz oscillator equation, the ellipsometry data of the test points close to the center of all the measurement points are analyzed in advance to obtain the inversion results; S3: Fitting the joint dispersion model, and using the evaluation function to determine the quality of the inversion results of the test points close to the center position, to obtain the parameters of the test points close to the center position; S4: performing ellipsometric data analysis on the test points on both sides of the test point close to the center position according to the evaluation function to obtain the best ellipsometric fitting result of the test point close to the center position; S5: After completing the ellipsometric data analysis of the test points on both sides, perform ellipsometric data analysis and evaluation on the test points adjacent to the test points on both sides that have not been analyzed, until the best fitting parameters including the film thickness and the dielectric function of all the ellipsometric measurement data points are obtained.

2. The method for fitting continuous gradient composition thin film ellipsometry measurement data according to claim 1, characterized in that: The joint dispersion model including the Drude equation and the Lorentz oscillator equation is: Among them, ε r is the real part of the dielectric function, ε i is the imaginary part of the dielectric function, λ is the wavelength of the incident light, P is the electric polarization intensity in the Drude dispersion equation, D is the inverse of the plasma wavelength in the Drude dispersion equation, E is the mean free path of electron motion in the Drude dispersion equation, A is the peak intensity of the Lorentz oscillator, L0 is the central wavelength of the Lorentz, and γ is the broadening of the Lorentz.

3. The method for fitting continuous gradient composition thin film ellipsometry data according to claim 2, characterized in that: The method for obtaining the inversion result in step S2 is: performing ellipsometric inversion using simulated annealing-ant colony algorithm, artificial ant colony algorithm or point-to-point algorithm to obtain the inversion result.

4. The method for fitting continuous gradient composition thin film ellipsometry data according to claim 3, characterized in that: The evaluation function is: Where n is the number of test points within the test range, m is the number of parameters involved in the fitting process, Ψ is the amplitude ratio and phase difference of the p-light before and after reflection, Δ is the amplitude ratio and phase difference of the s-light before and after reflection, and the subscripts cal and exp represent the calculated value and experimental value of the ellipsometric angle, respectively.

5. The method for fitting continuous gradient composition thin film ellipsometry measurement data according to any one of claims 2 to 4, characterized in that: The method for fitting the joint dispersion model in step S3 is: fitting the fitting parameters of the joint dispersion model using a simulated annealing algorithm.

6. The method for fitting continuous gradient composition thin film ellipsometric data according to claim 5, characterized in that: The fitting parameters are the electric polarization intensity P, the inverse of the plasma wavelength D, the mean free path E of the electron movement, the peak intensity A of the Lorentz oscillator, the central wavelength L0 of the Lorentz and the broadening γ of the Lorentz.

7. The method for fitting continuous gradient composition thin film ellipsometric data according to claim 6, characterized in that: The method of using the evaluation function to evaluate the quality of the inversion result in step S3 is: Input the inversion result into the evaluation function. If the output of the evaluation function is in the order of magnitude of 10 -4 If the magnitude of the evaluation function output result is greater than 10 -4 If the inversion result is not satisfactory, it is necessary to repeat step S2 to obtain the inversion result again.

8. The method for fitting continuous gradient composition thin film ellipsometry data according to claim 7, characterized in that: The test point parameters in step S3 are the fitting parameters, film thickness value and dielectric function of the test point close to the center position.

9. The method for fitting continuous gradient composition thin film ellipsometry data according to claim 7 or 8, characterized in that: The method for performing ellipsometric data analysis on the test points on both sides of the test point close to the center position in step S4 is: S41: performing ellipsometry data analysis using test points on both sides of a joint dispersion model including the Drude equation and the Lorentz oscillator equation, and setting the joint dispersion model including the Drude equation and the Lorentz oscillator equation and the film thickness value; S42: Obtain the inversion results of the two side points through ellipsometry data analysis, input the inversion results of the two side points into the evaluation function for evaluation, and if the output result of the evaluation function is within the order of magnitude of 10 -4 If the magnitude of the evaluation function output result is greater than 10 -4 If the inversion result is not satisfactory, the fitting parameters need to be adjusted and the ellipsometric data analysis needs to be repeated.

10. The method for fitting continuous gradient composition thin film ellipsometry measurement data according to claim 9, characterized in that: step The method for setting the joint dispersion model including the Drude equation and the Lorentz oscillator equation and the film thickness value described in S41 is: using the best ellipsometric fitting result as the initial value to set the electric polarization intensity P in the Drude dispersion equation, the inverse of the plasma wavelength D in the Drude dispersion equation, the mean free path E of electron motion in the Drude dispersion equation, the peak intensity A of the Lorentz oscillator, the central wavelength L0 of the Lorentz, the broadening γ of the Lorentz and the film thickness value in the joint dispersion model including the Drude equation and the Lorentz oscillator equation.

Images