JP2006220525A - Film thickness measurement method and film thickness measurement program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a film thickness measurement method and a film thickness measurement program for accurately finding a film thickness value in view of the effect of an inclination of a film surface and the thickness of a substrate, and not presupposing that the refractive index of the film is already-known. <P>SOLUTION: A spectral reflectivity model is prepared in view of the effect of the inclination of the film surface and the thickness of the substrate to perform a comparison operation of this model with measured spectral reflectivity, thereby accurately finding the film thickness value. First-stage and second-stage theoretical curves of reflectivity are prepared from a wavelength dispersion expression model of a refractive index. A film thickness value, which minimizes the mean square error between the second-stage theoretical curve and data on the measured spectral reflectivity, is used as a measurement result on a film thickness to be found. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a film thickness measurement method for obtaining a film thickness of a thin film formed on a substrate such as a glass substrate for a liquid crystal display device in a non-contact manner.

  Spectral interference non-contact film thickness measurement uses an interference phenomenon that is caused by the phase difference between the reflected light from the film surface and the reflected light from the boundary surface between the film and the substrate. It is. It is known that the intensity of the interference light depends on factors such as the wavelength of light applied to the film, the thickness of the film, the refractive index of the film, and the refractive index of the substrate.

  In the conventional film thickness measurement method, the measured spectral reflectance is compared with the theoretical spectral reflectance when the above factor is assumed to be a known value. For example, the measured spectral reflectance is compared with the theoretical spectral reflectance. The film thickness of the thin film as the object to be measured is determined based on the result of the mean square error of the reflectance.

  The theoretical spectral reflectance used in the above method is based on the premise that the surface of the object to be measured is completely flat, and further the reflected light from the object to be measured or the reflected light from the back surface of the substrate. It is calculated on the assumption that all light is guided to the light receiving element.

  However, when the thin film of the object to be measured is tilted (FIG. 1B) or depending on the thickness of the substrate (FIG. 1C), the reflected light from the object to be measured reaches the light receiving element. Since the amount of light is reduced, the measured spectral reflectance is smaller than the theoretical spectral reflectance. As a result, the mean square error between the measured spectral reflectance and the theoretical spectral reflectance is far from the true film thickness value. There have been problems of minimizing and mean square error not converging (no solution is found).

  In particular, this problem becomes significant in the microspectroscopic method in which light is irradiated to a minute region of the object to be measured through the objective lens.

  In recent years, miniaturization of a part to be measured has been desired in a manufacturing process of a member for a liquid crystal display device, a semiconductor element, and the like. For example, non-destructive and non-contact measurement of resist film thickness within the patterned pixels of a color filter, which is a member of a liquid crystal display device, is strongly desired, and the film surface after the polishing process is also flattened for the measurement target. Some have been performed (FIG. 2A), while others are pre-baked and the film surface is inclined (FIG. 2B). For a film having a flat film surface, a measurement method as in Patent Document 1 is known.

  As described above, since it is difficult to measure the film thickness of such an object with a spectral interference type film thickness measurement method, a part of the film is scraped off with a sharp blade so that the film surface and the substrate surface Measurements are made using a stylus type step gauge, but this method is cumbersome to prepare for measurement, takes time, and is a destructive measurement that destroys the object being measured. Can't expect to reduce losses.

  In addition, a method called a peak valley method is known as a method for measuring a film thickness by a spectral interference method, but since this method is based on the premise that the refractive index at each wavelength is known, It is necessary to measure the refractive index with an ellipsometer or the like in advance.

  However, the types of color resists differ depending on the application of the liquid crystal display device, and the types of color resists are not only diverse, but even the same type of color resist has a different refractive index before and after baking, so the refractive index is known. The application of the peak valley method on the premise of this is limited in terms of operation.

JP 2003-42722 A

  The present invention has been made in view of the above problems, and the problem is that a spectral reflectance model is created in consideration of the influence of the inclination of the film surface and the thickness of the substrate. A film thickness measurement method and a film thickness measurement program that do not assume that the refractive index of the film is known. It is in.

In order to solve the above-mentioned problem, in the invention according to claim 1, the sample surface on which the thin film is formed is irradiated with light, the reflected light from the sample is dispersed to obtain spectral reflectance data, In the film thickness measurement method for measuring the film thickness of the thin film by comparing the measured spectral reflectance with a theoretical curve obtained by theoretical calculation,
Using the refractive index of the substrate and the estimated value of the refractive index of the thin film, determining the interference order from the wavelength at which the measured spectral reflectance is maximized and minimized;
Using the determined interference order, obtaining a film thickness value corresponding to the maximum and minimum wavelengths;
There is provided a film thickness measuring method including a step of specifying a film thickness range that the thin film can take from a value based on the film thickness value.

Further, in the invention according to claim 2, the sample surface on which the thin film is formed on the substrate is irradiated with light, the reflected light from the sample is dispersed to obtain spectral reflectance data, and the measured spectral reflectance is obtained. In the film thickness measurement method for measuring the film thickness of the thin film by comparing with a theoretical curve obtained by theoretical calculation,
Using the refractive index of the substrate and the estimated value of the refractive index of the thin film, determining the interference order from the wavelength at which the measured spectral reflectance is maximized and minimized;
Using the determined interference order, obtaining a film thickness value corresponding to the maximum and minimum wavelengths;
After performing the step of specifying the film thickness range that the thin film can take from the value based on the film thickness value,
Re-determining the refractive index of the thin film from the maximum and minimum wavelengths of the measured spectral reflectance for each film thickness value by using the interference order and changing the film thickness value;
Creating a chromatic dispersion model of the refractive index of the thin film from the refractive index corresponding to the maximum and minimum wavelengths determined for each film thickness value;
The film according to claim 1, further comprising: creating a first-stage theoretical curve of reflectivity based on each film thickness value and the chromatic dispersion model determined for each film thickness value. A thickness measurement method is provided.

  Further, in the invention according to claim 3, the step of creating the wavelength dispersion type model includes a step of selecting according to the number of refractive indexes corresponding to the maximum and minimum wavelengths. The described film thickness measuring method is provided.

  The invention according to claim 4 is characterized in that the first-stage theoretical curve is a linear combination of reflection derived from the thin film and back-surface reflection of the substrate. A film thickness measuring method is provided.

  In the invention according to claim 5, the value based on the reflectance at the minimum wavelength of the theoretical curve of the first stage is compared with the value based on the reflectance at the minimum wavelength of the measured spectral reflectance, and the comparison calculation is performed. And calculating the offset amount, and creating a second stage theoretical curve obtained by offsetting the first stage theoretical curve by the amount of the offset, the second stage theoretical curve, and the actual measurement 5. The film thickness measurement method according to claim 2, wherein the film thickness value when the mean square error of the spectral reflectance data is minimized is a measurement result of the film thickness to be obtained. .

Further, the invention according to claim 6 includes a step of judging validity of a refractive index corresponding to a wavelength at which the measured spectral reflectance is maximized / minimized, which is used when obtaining the measurement result of the film thickness. ,
6. The film according to claim 5, wherein when the refractive index is determined to be appropriate, a film thickness value of the film thickness measurement result is displayed as a measurement result, and when the refractive index is determined to be abnormal, an error is displayed. A thickness measurement method is provided.

  According to a seventh aspect of the invention, there is provided a film thickness measurement program that causes a computer to perform the film thickness measurement method according to any one of the first to sixth aspects.

  According to the present invention, a spectral reflectance model that takes into account the influence of the inclination of the film surface and the thickness of the substrate is created, and the film thickness value is obtained accurately by performing a comparison operation between this model and the measured spectral reflectance. Can do. Moreover, the film thickness value can be obtained without assuming that the refractive index of the film is known.

  First, an outline of the film thickness measuring method of the present invention will be described.

  FIG. 3 is a flowchart showing the procedure of the film thickness measuring method of the present invention. Hereinafter, each step will be described in detail.

In step (hereinafter sometimes abbreviated as “S”) 1, an actual spectral reflectance (hereinafter referred to as “R _Exp (λ)”) is obtained.

が成り立つ。ここで、ｎ_Ｆ（λ_{Ｖａｌｌｅｙ}）、ｎ_Ｆ（λ_Ｐｅａｋ）は、膜の屈折率ｎ_Ｆの波長分散を考慮したもので、ｎ_Ｆが波長λの関数であることを表している。 In S2, an interference order m, which will be described later, and an initial value (<d> -α) are determined from the maximum and minimum wavelengths of R _Exp (λ). FIG. 4 shows R _Exp (λ) in the blue cell of the color filter obtained by the normal microspectroscopic method. Assume that the refractive indices of air, film, and substrate are 1.0, n _F , and n _S (where n _F > n _S > 1.0 and n _S are known in the measurement wavelength range). Since the optical path difference between the reflected light R _1F from the film surface and the reflected light R _FS from the boundary surface between the film and the substrate is 2n _F d and n _F > n _S > 1.0, R _1F and R The phase of _FS changes by π compared to the phase of incident light. Therefore, assuming that the minimum wavelength is λ _Valley , the maximum wavelength is λ _Peak , the film thickness is d, and the interference order is m (integer),

Holds. Here, n _F (λ _Valley ) and n _F (λ _Peak ) take into consideration the wavelength dispersion of the refractive index n _F of the film, and represent that n _F is a function of the wavelength λ.

In S2, assuming that n _F (λ _Valley ) and n _F (λ _Peak ) of Equations 1 and 2 are 1.7, the interference order m and the initial value (<d> −α) of the film thickness are obtained. Note that n _F (λ _Valley ) and n _F (λ _Peak ) are assumed to be 1.7 because the refractive index n _F of a general color resist is the refractive index n _S (≈1. This is because it is slightly larger than 5). That is, at this stage, n _F may be a value slightly larger than n _S , and strictness is not required.

なので、適当な干渉次数ｍを数３、数４に代入すれば、Ｒ_Ｅｘｐ（λ）の各極大・極小波長に対応した複数の膜厚値が算出されることになる。 Assuming that n _F (λ _Valley ) and n _F (λ _Peak ) of Equations 1 and 2 are 1.7,

Therefore, if an appropriate interference order m is substituted into Equations 3 and 4, a plurality of film thickness values corresponding to the maximum and minimum wavelengths of R _Exp (λ) are calculated.

Hereinafter, a method for determining the interference order m and the initial value (<d> −α) of the film thickness will be described.
(I) Wavelengths λ ₁ , λ ₂ , λ ₃ ,... Taking all local maximum values and local minimum values are obtained from R _Exp (λ) (FIG. 4).
(Ii) If the shortest wavelength gives the minimum value among the wavelengths giving the maximum value and the minimum value obtained in (i), an integer m is given as a temporary interference order, and the shortest wavelength has the maximum value. If given, (m + 0.5) is given as a temporary interference order. Since the maximum value and the minimum value appear alternately, the value obtained by subtracting 0.5 may be used as the interference order for the wavelength that gives the remaining maximum value and minimum value (hereinafter, this series of interference orders is expressed as “ Interference order set ”). In this way, the film thicknesses d ₁ , d ₂ , d ₃ ,... Are obtained for λ ₁ , λ ₂ , λ ₃ ,.
(Iii) Integers (m + 1), (m + 2), (m + 3),... Are given as temporary interference orders, and the same procedure as in (ii) is performed.
(Iv) The difference between the maximum value and the minimum value of d ₁ , d ₂ , d ₃ ,... Obtained by giving each interference order set is calculated, and the interference order set when this becomes the minimum is obtained.
(V) An average value <d> of d ₁ , d ₂ , d ₃ ,... With respect to the interference order set obtained in (iv) is obtained, and <d> −α is obtained as an initial value of the film thickness.

  Here, α determines the range in which d is shaken when d is treated as a variable in step S3 and subsequent steps, and can be arbitrarily set by the user. That is, after S3, d is shaken in the range of (<d> −α) ≦ d ≦ (<d> + α).

In S4, m determined in S2 is fixed, and n corresponding to the variable d is obtained again. A series of refractive indexes n ₁ , n ₂ , n ₃ ,... Corresponding to the wavelengths λ ₁ , λ ₂ , λ ₃ ,.

Subsequently, in order to obtain the refractive index in the entire measurement wavelength range, the wavelength dispersion formula n (λ) of the refractive index is determined using n ₁ , n ₂ , n ₃ ,. . As a model of n (λ), a known model equation may be used. As can be seen from FIG. 5 showing the spectral transmittance of various color resists, in the case of color resists, a transparent film having no absorption in the near infrared region is used. Since it can be handled, it is preferable to select the near infrared region as the measurement wavelength region and determine n (λ) using the Sellmeier model or the Couchy model.

S4 may include a step of selecting a model expression of n (λ) based on the number of n ₁ , n ₂ , n ₃ ,.

Here, a method of determining a wavelength dispersion formula (hereinafter referred to as “n _Cellmeier (λ)”) based on the _Sellmeier model and a wavelength dispersion formula (hereinafter referred to as “n _Cauchy (λ)”) based on the Cauchy model will be described. . n _Sellmeier (λ) is expressed in the form of _Equation 7, and is _{determined by obtaining} coefficients a and b that are closest to n ₁ , n ₂ , n _{3 ,.} Further, n _Cauchy (λ) is expressed in the form of _Formula 8, and is determined by obtaining coefficients A, B, and C that are closest to n ₁ , n ₂ , n ₃ ,. The coefficients a and b in Expression 7 and the coefficients A, B, and C in Expression 8 can be easily obtained by using a known nonlinear least square method (for example, Levenberg-Marquardt method) or a linear least square method.

As described above, the number of n ₁ , n ₂ , n ₃ ,... Corresponding to λ ₁ , λ ₂ , λ ₃ ,... _{Serves as} a selection criterion for n _Sellmeier (λ) and n _Cauchy (λ). , N ₁ , n ₂ , n ₃ ,..., N _Cauchy (λ) including three coefficients cannot be determined, so n _Sellmeier (λ) is selected, and n ₁ , When the number of n ₂ , n ₃ ,... is 3 or more, n _Cauchy (λ) may be selected. If n ₁ , n ₂ , n ₃ ,... Is 1 or less, it is impossible to determine n _Sellmeier (λ). Since it is 3000 nm, if a spectroscope having a measurement wavelength range of about 900 to 1600 nm is used, the number of n ₁ , n ₂ , n ₃ ,..., That is, the maximum and minimum values of R _Exp (λ) are obtained. It is possible to avoid a situation in which the number of wavelengths λ ₁ , λ ₂ , λ ₃ ,.

In S5, using the variable d and n (λ) corresponding to the variable d obtained in S4, a first-stage theoretical curve expressed by Equation 9 (hereinafter referred to as “R ′ _Theory (λ)”). Create Here, P and Q in Equation 9 are parameters that can be arbitrarily set by the user, and the optimum values can be experimentally obtained in advance from various samples. R ₁ (λ) is a spectral reflectance calculated theoretically on the premise that light is perpendicularly incident on the single-layer film, and is expressed as in Expression 10. R ₂ (λ) is the back surface reflection of the substrate and is expressed by Equation 11. That is, Equation 9 is a linear combination of the spectral reflectance R ₁ (λ) of the so-called single-layer film model and the back surface reflection R ₂ (λ). Note that r _1F and r _{FS in Equation} 10 are the Fresnel reflection coefficient at the film surface and the Fresnel reflection coefficient at the interface between the film and the substrate, and are expressed by Equations 12 and 13, respectively, and R ₀ (λ) in Equation 11 Is the square of r _1F .

In S6, a second-stage theoretical curve (hereinafter referred to as “R _Theory (λ)”) in which the minimum values of R _Exp (λ) and R ′ _Theory (λ) are aligned is created. Hereinafter, a method of creating R _Theory (λ) will be described.

The wavelength at which the R _Exp (lambda) is minimum _{λ Valley1, λ Valley2, λ Valley3} , ..., R'Theory (λ) is Ramuda' a wavelength at which the minimum _{Valley1, λ'Valley2, λ'Valley3,} When ... to _the minimum value of _{R Exp} (lambda) is _{R Exp (λ Valley1), R} Exp (λ Valley2), R Exp (λ Valley3), ..., R' minimum value of _{Thory (λ)} is _R'Theory ( _λ ′ _{Valley 1} ), R ′ _Theory ( _λ ′ _{Valley 2} ), R ′ _Theory (λ ′ _{Valley 3} ),..., and therefore, the average value of the minimum values of R _Exp (λ) (hereinafter, “<R _Exp (λ _Valley)> and _{"hereinafter),} the average value of the minimum value of the _{R'Theory (λ)} (hereinafter referred to as"_{<R'Theory(λ' alley)>} referred to as ") _is the number of the minimum value of _{R Exp} (lambda) _I, the number of local minima of _R'Theory (lambda) as J, number 14, obtained by the number 15.

Therefore, in order to align the minimum values of R _Exp (λ) and R ′ _Theory (λ), R ′ _Theory (λ) is set to <R _Exp (λ _Valley )> and <R ′ _Theory (λ ′ _Valley )>. What is necessary is just to offset by the offset amount represented by the difference. That is, R _Theory (λ) is expressed by Equation 16.

In S7, the mean square error of R _Exp (λ) and R _Theory (λ) is calculated.

  Thereafter, in S8, the value of the variable d is increased by δd, and the processing of S4 to S7 is performed for each d. Note that the mean square error at each d is sequentially stored in the memory in association with the value of d. In this way, when the process up to substituting <d> + α for d is completed, S9 becomes “Yes” and the process proceeds to S10, and the film thickness when the mean square error is minimized (hereinafter referred to as “thickness”). (Denoted “D”).

In S11, when d = D, the validity of a series of refractive indexes n ₁ , n ₂ , n ₃ ,. The determination criterion may be whether n ₁ , n ₂ , n ₃ ,... Is a value slightly larger than the refractive index n _S (≈1.5) of the glass substrate. Thus, when it is determined that n ₁ , n ₂ , n ₃ ,... Are normal values, the process proceeds to S 13 to display D as a measurement result on the monitor, while n ₁ , n ₂ , n _3. If it is determined that the values are abnormal values, the process proceeds to S12 to display an error and prompt the user to remeasure.

  According to the procedure shown in FIG. 3, a spectral reflectance model that takes into account the influence of the inclination of the film surface and the thickness of the substrate is created, and the film thickness value is accurately obtained by performing a comparison operation between this model and the measured spectral reflectance. Can be sought. Moreover, the film thickness value can be obtained without assuming that the refractive index of the film is known. Each step described above can be automatically executed by creating a computer program for appropriately inputting necessary parameter values and physical property values.

  In the present invention, as described above, when the object to be measured is a color filter having a flattened film surface after the polishing step, or a color filter in which the film surface only prebaked is inclined, Alternatively, by setting P and Q of Equation 9 to appropriate values obtained in advance according to the thickness of the glass substrate, for example, a stage as shown in FIG. Non-destructive measurement at is possible. In other words, it is possible to detect a defect factor of a product in a non-contact / non-destructive manner and at an early stage, so that a reduction in line loss and an improvement in yield can be expected.

  Next, examples of the film thickness measuring method of the present invention will be described.

FIG. 4 shows R _Exp (λ) in the blue cell of the color filter obtained by the normal microspectroscopic method. Wavelengths λ ₁ , λ ₂ , λ ₃ ,... Taking all local maxima and minima are obtained from FIG. 4, and film thicknesses d ₁ , d ₂ , d ₃ , ... _shows d _1, d 2, _{d 3,} the difference between ... maximum value and the minimum value of _the result of obtaining d _1, d 2, _{d 3,} ... average of <d> Table 1.

From Table 1, the interference order set (case 2 in the table) when the difference between the maximum value and the minimum value of d ₁ , d ₂ , d ₃ ,... Is minimum and <d> = 1734.64 are determined. It was.

  In S3, a number <d> -α based on <d> obtained in S2 is set as an initial value of d. Here, α determines the range in which d is shaken when d is treated as a variable in step S3 and subsequent steps, and can be arbitrarily set by the user. That is, after S3, d is shaken in the range of (<d> −α) ≦ d ≦ (<d> + α). From experience based on various experiments by the inventors, it has been confirmed that if α is set to about 200 nm, d = D at which the mean square error is minimized can be extracted in S10.

In S4, m determined in S2 is fixed, and n corresponding to the variable d is obtained again. A series of refractive indexes n ₁ , n ₂ , n ₃ ,... Corresponding to the wavelengths λ ₁ , λ ₂ , λ ₃ ,. In order to obtain the rate, the wavelength dispersion formula n (λ) of the refractive index is determined using the n ₁ , n ₂ , n ₃ ,. FIG. 6 shows n ₁ , n ₂ , n ₃ ,..., N _Sellmeier (λ), n _Cauchy when using the interference order set of Case 2 in Table 1 and d = <d> _−α≈1535 nm. (Λ) is shown (α = 200 nm). At this time, the total number of n ₁ , n ₂ , n ₃ ,... Is 6, and n _Cauchy (λ) shows a better approximation to n ₁ , n ₂ , n _{3 ,.} Therefore, n _Cauchy (λ) is selected as a model of n (λ) and the process _proceeds to the next step.

Since the variable d and n (λ) obtained from the variable d are determined, it becomes possible to create R ′ _Theory (λ) in S5 and R _Theory (λ) in S6. Note that P and Q in Equation 9 are parameters that can be arbitrarily set by the user. However, when the object to be measured is a color filter having a flattened film surface after the polishing process, the film surface is only pre-baked. Is an inclined color filter, or it is important to set an appropriate value determined in advance according to the thickness of the glass substrate or the like.

In subsequent S7, it becomes possible to calculate the mean square error of R _Exp (λ) and R _Theory (λ). FIG. 7 shows that R _Exp (λ) and P and Q in _Equation 9 are P = 0.95 and Q = 0.70 as optimum values when the film surface is flat, and d = <d> −α≈. FIG. 7 shows R _Theory (λ) obtained using 1 _Cauchy (λ) in FIG. 6 at 1535 nm.

Thereafter, in S8, the value of the variable d is increased by δd, the processing of S4 to S7 is performed for each d, and when the processing up to substituting <d> + α is completed, S9 becomes “Yes”. The process proceeds to S10, and the film thickness D when the mean square error is minimized is extracted. FIG. 8 shows the relationship between d and the mean square error of R _Exp (λ) and R _Theory (λ) when d is swung in the range of (<d> −α) ≦ d ≦ (<d> + α). Is shown. In this embodiment, δd = 5 nm, but it is needless to say that the value of δd can be set smaller or larger.

From FIG. 8, the film thickness D when the mean square error between R _Exp (λ) and R _Theory (λ) is minimized was determined to be 1790 nm. Note that FIG. 9 is _n 1 in _{d = D, n 2, n} 3, ..., n Sellmeier (λ), but shows the _{n Cauchy (λ), n 1} , n 2, n 3, ... is Since the value is slightly larger than the refractive index n _S (≈1.5) of the glass substrate and is determined to be a normal value in S11, D is displayed on the monitor as a measurement result. Further, FIG. 10 is obtained by using R _Exp (λ) and P and Q of Equation 9 with P = 0.95 and Q = 0.70, d = D, and n _Cauchy (λ) of FIG. R _Theory (λ) is shown.

  In the same procedure, the red cell and the green cell of the color filter were measured, and D = 1815 nm was obtained in the red cell and D = 1835 nm was obtained in the green cell.

In order to confirm the validity of the above measurement results, the film thicknesses in the red cell, green cell, and blue cell were measured with a stylus type step meter, as shown in Table 2, all of which are the present invention. Thus, it was confirmed that an accurate film thickness value was obtained.

  When the same procedure was performed for the green cell shown in FIG. 2B with P and Q of Equation 9 set to P = 0.95 and Q = 0.70, D = 2750 nm was obtained. The measurement result by a stylus type step gauge was 2710 nm, and the difference was larger than that when the film surface was flattened after the polishing step (FIG. 2A).

  Here, looking at the profile shape of the green cell in FIG. 2B, the amount of reflected light reaching the light receiving element is small because the film surface is inclined as shown in FIG. 1B. It was expected to become. The reason why the film surface of the green cell is inclined is that the color resist flows into the green cell from the red + green cell shown in FIG. 2B. Accordingly, the coating conditions of the color resist coating apparatus are set in consideration of the amount flowing from the adjacent cell. For these reasons, there is an increasing demand for measuring the film thickness in a non-contact and non-destructive manner immediately after the color resist is applied.

Therefore, when P and Q in Equation 9 are set to P = 0.80 and Q = 0.60 as optimum values when the film surface is inclined, D = 2710 nm, which is measured by a stylus type step meter. A value consistent with the results was obtained.

The figure which shows the difference in the mode of reflection of reflected light by the state of a thin film, and the thickness of a board | substrate. The figure which shows the measurement result by a stylus type level difference meter. The flowchart which shows the procedure of the film thickness measuring method of this invention. The figure which shows _RExp ((lambda)) in the blue cell of a color filter. The figure which shows the spectral transmittance of various color resists. The figure which displayed the example of a wavelength dispersion type model of a refractive index. The figure which shows the theoretical curve of a 2nd step. R _Exp (lambda) and shows the relationship between the mean square error of _{R Theory (λ).} The figure of a wavelength dispersion type in d = D. The figure which shows _RTheory ((lambda)) calculated | required as d = D.

Claims (7)

Irradiate the surface of the sample with a thin film on the substrate, divide the reflected light from the sample to obtain spectral reflectance data, and compare the measured spectral reflectance with the theoretical curve obtained by theoretical calculation. In the film thickness measuring method for measuring the film thickness of the thin film,
Using the refractive index of the substrate and the estimated value of the refractive index of the thin film, determining the interference order from the wavelength at which the measured spectral reflectance is maximized and minimized;
Using the determined interference order, obtaining a film thickness value corresponding to the maximum and minimum wavelengths;
A method of measuring a film thickness comprising the step of specifying a film thickness range that the thin film can take from a value based on the film thickness value. Light is applied to the surface of the sample on which a thin film is formed on the substrate, the reflected light from the sample is dispersed to obtain spectral reflectance data, and the measured spectral reflectance is compared with a theoretical curve obtained by theoretical calculation. In the film thickness measuring method for measuring the film thickness of the thin film,
Using the refractive index of the substrate and the estimated value of the refractive index of the thin film, determining the interference order from the wavelength at which the measured spectral reflectance is maximized and minimized;
Using the determined interference order, obtaining a film thickness value corresponding to the maximum and minimum wavelengths;
After performing the step of specifying the film thickness range that the thin film can take from the value based on the film thickness value,
Re-determining the refractive index of the thin film from the maximum and minimum wavelengths of the measured spectral reflectance for each film thickness value by using the interference order and changing the film thickness value;
Creating a chromatic dispersion model of the refractive index of the thin film from the refractive index corresponding to the maximum and minimum wavelengths determined for each film thickness value;
The film according to claim 1, further comprising: creating a first-stage theoretical curve of reflectivity based on each film thickness value and the wavelength dispersion formula model determined for each film thickness value. Thickness measurement method.   3. The film thickness measuring method according to claim 2, wherein the step of creating the wavelength dispersion model includes a step selected according to the number of refractive indexes corresponding to maximum and minimum wavelengths.   4. The film thickness measuring method according to claim 2, wherein the first-stage theoretical curve is a linear combination of a reflection derived from the thin film and a back surface reflection of the substrate.   Comparing the value based on the reflectance at the minimum wavelength of the theoretical curve of the first stage with the value based on the reflectance at the minimum wavelength of the measured spectral reflectance, and calculating the offset amount by the comparison operation; Creating a second stage theoretical curve obtained by offsetting the first stage theoretical curve by the offset amount, and the mean square error between the second stage theoretical curve and the measured spectral reflectance data is minimized. The film thickness measurement method according to any one of claims 2 to 4, wherein the film thickness value at the time is a measurement result of the film thickness to be obtained. Determining the appropriateness of the refractive index corresponding to the wavelength at which the measured spectral reflectance is maximized / minimum used when obtaining the measurement result of the film thickness;
6. The film according to claim 5, wherein when the refractive index is determined to be appropriate, a film thickness value of the film thickness measurement result is displayed as a measurement result, and when the refractive index is determined to be abnormal, an error is displayed. Thickness measurement method.   A film thickness measurement program for causing a computer to perform the film thickness measurement method according to claim 1.

Images