JPH1151618A - Film thickness detecting method for multi-layered thin film - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the improved film thickness measuring instrument which can measure a thicker multi-layered film with high precision. SOLUTION: A sample 10 having a multi-layered thin film is irradiated with the light from a white light source 1 and the light reflected by the sample is spectrally diffused by a multichannel spectroscope 8. Further, a signal processor 7 after reducing reflection spectra other than the object of measurement according to the spectrum patterns of the light source 1 and the respective layers of the sample 10 performs the wave front conversion of the reflection spectrum. Then a DC component is removed through a smoothing differentiating process and then fast Fourier transformation is performed to detect the film thickness.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting the thickness of a multilayer thin film, and more particularly, to a method for detecting the thickness of a multilayer thin film such as an organic photoconductor layer of an electrophotographic photosensitive member. It relates to a detection method.

[0002]

2. Description of the Related Art The repetitive reflection interferometry, in which the thickness of a light-transmitting thin film is measured by reflected light, is excellent in that the sample can be measured nondestructively. The measurement method using a visible light of the multi-channel detector system detected by the method is advantageous in that a measuring device without a mechanical driving unit can be used. Such a measuring method analyzes the white reflected light spectrum with a multi-channel spectroscope, calculates the reflection intensity from the obtained amplitude reflectance, and then, based on the wavelength and the refractive index of the material giving the extreme value among the wavelengths, An approximate value of the film thickness is obtained by a repeated method using the Newton method and setting the film thickness as an initial value.

[0003] However, the above-mentioned method for measuring a thin film is as follows.
When applied to a multilayer film such as an organic photoconductor layer of an electrophotographic photosensitive member, a problem occurs in signal processing. That is, in the above-mentioned iterative method, when the value of the film thickness set as the initial value of the calculation is different from the actual value, the solution may not be obtained quickly or may be a different solution. . In addition, when the number of film layers is large, the number of repetitions increases, and the processing time becomes long, which is not suitable for practical processing such as in-line.

Therefore, as described in Japanese Patent Application Laid-Open No. Hei 7-294220, the present inventors first used a fast Fourier transform (FFT) technique to increase the thickness of a multilayer thin film at high speed and high speed. A film thickness detection method that accurately detects the thickness is proposed. The method for detecting the thickness of a multilayer thin film described in the above publication irradiates a sample with white light, disperses light reflected from the sample, performs a Hanning window process on the reflection spectrum, and performs fast Fourier transform to obtain an energy spectrum. After that, the waveform is subjected to wave number conversion processing to specify the thickness of the thin film.

[0005]

Incidentally, in recent electrophotographic photoreceptors, in order to cope with color copying and higher resolution image processing, the thickness of the electrophotographic photoreceptor is more uniform and 0.4 mm.
A thinner film of about 1.5 μm is formed. As a result, in the above-described film thickness detection method, there is a case where sufficient analysis accuracy cannot be obtained due to the influence of the reflected light including the spectrum of the base on the reflected light from the thin film layer to be measured. The present invention
In view of such circumstances, an object of the present invention is to provide an improved film thickness detection method capable of measuring a thinner multilayer film with higher accuracy.

[0006]

In order to solve the above-mentioned problems, a method for detecting the thickness of a multilayer thin film according to the present invention includes irradiating a sample having a multilayer thin film with white light and spectrally reflecting light reflected from the sample. In addition, based on the light source and the spectral pattern of each layer, after reducing the reflection spectrum outside the measurement target, the reflection spectrum is wavenumber-converted, then smoothed and differentiated to remove the DC component, and then subjected to fast Fourier transform. Thus, the thickness of the thin film is detected.

In the above-described film thickness detection method, in order to effectively remove a DC component in a frequency range, it is preferable to perform a smoothing differentiation process using a high-pass filter (y _t ) represented by the following equation. .

[0008]

(Equation 4)

Further, in order to make effective peaks appear in the frequency range and to efficiently perform the fast Fourier transform, it is preferable to perform a fast Fourier transform by performing a process using a Hanning window (W _n ) represented by the following equation.

[0010]

(Equation 5)

Further, in order to make the effective peaks in the frequency region more apparent, a process using a Slepian window, which is an eigenvector of the matrix (A) represented by the following equation, is performed instead of the Hanning window, and a fast Fourier transform is performed. Good.

[0012]

(Equation 6)

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a method for detecting the thickness of a multilayer thin film according to the present invention will be described with reference to the drawings. FIG. 1 is a block diagram illustrating a configuration example of a film thickness detection device. FIG.
FIG. 3 is a cross-sectional view schematically illustrating a layer configuration of a sample. FIG.
FIG. 4 is a diagram showing a frequency response when a high-pass filter is used for smoothing differentiation processing. FIG. 4 is a diagram illustrating a processing example using a Hanning window used in the processing of the frequency range. FIG. 5 is a diagram illustrating a processing example using a Slepian window used in the processing of the frequency domain.

The method for detecting a film thickness according to the present invention uses a film thickness detecting device as shown in FIG.
Is a method for non-destructively detecting the thickness of a predetermined film. Examples of the sample (10) having a multilayer thin film include an electrophotographic photoreceptor. The electrophotographic photoreceptor includes, for example, an aluminum conductive layer formed on the surface of a sheet-like or cylindrical substrate, and a photosensitive layer such as a charge generation layer and a charge transfer layer formed on the conductive layer via an undercoat layer. It is formed sequentially.

The undercoat layer is a coating film of nylon or the like formed with a thickness of about 0.5 to 3 μm. The photosensitive layer is formed by a conventionally known coating solution of various photoconductor materials. As such a coating liquid, various conventionally known liquids composed of a photoreceptor material and one or more solvents can be used. As the charge generating substance, other than azo pigments such as Sudan Red and Diane Blue, disazo pigments, quinone pigments,
Examples include phthalocyanine pigments, pyrylium salts, azurenium salts and the like. Examples of the charge transport material include compounds having a skeleton of a polyaromatic compound such as anthracene or pyrene or a nitrogen-containing cyclic compound such as indole or carbazole in a main chain or a side chain, and a hole transport material such as a hydrazone compound. . When the photosensitive layer is a stacked type, for example, the charge generation layer is formed with a thickness of 0.3 to 1.0 μm, and the charge transfer layer is formed with a thickness of 15 to 40 μm.

In the following description, the sample (1
The layer configuration 0) will be described as a layer configuration as shown in FIG.
In the layer configuration shown in FIG. 2, the symbols (d ₁ , d ₂ ... D _N )
Indicates the thickness of each thin film layer on the surface of the base material, and the symbols (n ₀ , n
₁ , n ₂ ... _{N N} ) indicate the refractive index of the substrate and each thin film layer on its surface, and the symbol (n _{N + 1} ) indicates the refractive index of air on the surface of the sample (10).

The film thickness detecting device shown in FIG. 1 mainly includes a white light source (1), optical fibers (2, 3), a multi-channel spectroscope (6), and a signal processing device (7). The optical fiber (2, 3) is a bifurcated fiber, and converts the light emitted from the white light source (1) into a sample (10).
An optical fiber (2) for irradiating the sample and an optical fiber (3) for capturing the reflected return light from the sample (10) and introducing it to the multi-channel spectrometer (6). The multi-channel spectrometer (6) includes a spectroscope (4) for splitting incident light from the optical fiber (3) and a multi-channel detector (5) for detecting the intensity of the spectrum and outputting a signal to the signal processing device (7). ). The signal processing device (7) includes an arithmetic processing unit that performs a predetermined arithmetic process based on the input signal.

In the film thickness detecting method of the present invention, as shown in FIG. 1, first, the light of the white light source (1) is perpendicularly applied to the sample (10) via the optical fiber (2) (θ = 0). And the light reflected by the sample (10) is introduced into the spectroscope (4) of the multi-channel spectroscope (6) through the optical fiber (3) to be separated. Next, the intensity of the spectral spectrum is detected by the multi-channel detector (5), and the obtained intensity signal of the spectral spectrum is input to the signal processing device (7) for arithmetic processing.

In the present invention, when a reflection spectrum is input to the signal processing device (7), first, each layer of the sample (10) is prevented in order to prevent an increase in measurement error due to the influence of the intensity due to absorption of each layer. In the above, a wavelength range where there is no absorption that transmits to the lower layer side is selected. Further, in order to complement the interference between the thin films forming each layer, the reflection spectra of the layers other than the object to be measured are reduced in the wavelength range based on the light source and the reflection spectrum pattern of each layer confirmed in advance. In particular, when the sample (10) is an electrophotographic photosensitive member, it is effective to correct the peak of the reflection spectrum due to aluminum constituting the base material and the conductive layer.

Then, the obtained reflection spectrum is converted into a signal in a wavelength range from a wavelength range by performing a wave number conversion. When the signal in the wavelength range is converted into the reciprocal of the wavelength (λ), that is, the wave number (σ), d = (1 / 2n) (1 / Δσ)
From the relationship (where n is the refractive index), the film thickness d is expressed by the following equation: Δσ = σ ₁ −σ ₂ (σ _i = λ _i ⁻¹ , i = 1,
2) only, and is independent of the measurement start wavelength (λ ₁ ).

[0021]

(Equation 7)

Therefore, in order to remove the DC component, the signal subjected to wave number conversion is subjected to smoothing differentiation processing. In the present invention,
In performing the smoothing differentiation process, a so-called high-pass filter (y _t ) represented by the following equation (Chbyshev High-Pass Filte)
r) is used. For example, a high-pass filter (y
_{According to t} ), in a state where a signal of 20 cycles or less is attenuated to about 40 db, a frequency response as shown in FIG. 3 is obtained. By using such a high-pass filter (y _t ), a DC component in a frequency domain can be effectively removed.

[0023]

(Equation 8)

Next, a DC component is removed by a high-pass filter (y _t ), and the wave number signal subjected to the smoothing differentiation processing is subjected to processing by a predetermined window function. As a window function,
So-called Boxcar window, Hanning window (
Hanning window, Slepian window
) May be used.

The reason why the above window function is applied to the wave number signal is as follows. That is, since the number of sensors of the multi-channel detector (5) for detecting the reflection spectrum is a finite number, for example, 1024, if the subsequent fast Fourier transform is directly performed, the values at both ends of the signal (the first and last sensors) are obtained. Is greatly influenced by the result.

In the present invention, it is preferable to perform processing using an improved Hanning window (W _n ) represented by the following equation. By such a Hanning window process, an effective peak in the frequency domain can be made obvious. For example, the magnitude of the amplitude when 1024 pieces of measurement data are processed by the Hanning window (W _n ) is corrected as shown in FIG.

[0027]

(Equation 9)

In the present invention, instead of the above Hanning window, a process using a Slepian window, which is an eigenvector of the matrix (A) represented by the following equation, may be performed. By such Slepian window processing, effective peaks in the frequency domain can be made more apparent. For example, the magnitude of the amplitude when 1024 pieces of measurement data similar to the above are processed by the Slepian window is corrected as shown in FIG.

[0029]

(Equation 10)

As a result of the processing using each window function as described above, it is possible to improve the separation of the frequency of the reflected light from the thin film layer to be measured, which is closer to the frequency data obtained in the following process. Then, the signal subjected to each window processing can be processed as thickness data by performing a fast Fourier transform,
The thickness d of a predetermined thin film can be calculated at high speed. Further, in performing the fast Fourier transform, in order to refine the basis in the fast Fourier transform and improve the conversion (approximation) accuracy, 0 is expanded after the obtained 1024 signals, for example, and the total number of signals is increased by N times. By doing so, the accuracy can be improved N times.

In the calculation of the film thickness d, first, the amplitude reflectance (R) approximated by the following equation is calculated from the data obtained by the fast Fourier transform, and the magnitude | R | ² of the intensity spectrum is obtained.

[0032]

[Equation 11]

Next, since the following equation holds for the j-th peak position, the known refractive index n _j of each layer is used and an initial film thickness d _jo (j = 1... N) is set. The thickness of each layer can be calculated by sequentially calculating from.

[0034]

(Equation 12)

As described above, in the film thickness detection method of the present invention, the reflection spectrum outside the object of measurement is reduced in the wavelength range, the DC component is removed by smoothing differentiation in the frequency range, and Because of the fast Fourier transform, for example, an extremely thin film thickness of 0.4 to 1.5 μm in a multilayer film can be detected at high speed and with extremely high accuracy. Further, when the smoothing differential processing is performed using the high-pass filter, the DC component in the frequency range can be effectively removed, so that the film thickness can be specified with higher accuracy. Then, in the case where fast Fourier transform is performed by performing processing using an improved Hanning window or Slepian window in the frequency domain,
Close frequencies can be separated with high accuracy, and the film thickness can be specified with higher accuracy.

[0036]

According to the method for detecting the thickness of a multilayer thin film according to the present invention, the reflection spectrum of a layer not to be measured is reduced in the wavelength range, and the DC component is removed by smoothing and differentiation in the frequency range. In addition, because of the fast Fourier transform, an extremely thin film thickness in the multilayer film can be detected at high speed and with extremely high accuracy.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration example of a film thickness detection device.

FIG. 2 is a cross-sectional view schematically illustrating a layer configuration of a sample.

FIG. 3 is a diagram illustrating a frequency response when a high-pass filter is used in the smoothing differentiation process.

FIG. 4 is a diagram showing an example of processing using a Hanning window used in processing in a frequency range.

FIG. 5 is a diagram showing a processing example using a Slepian window used in processing in a frequency domain;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 White light source 2 Optical fiber 3 Optical fiber 4 Spectroscope 5 Multi-channel detector 6 Multi-channel spectrometer 7 Signal processing device

Claims (4)

[Claims] 1. A sample having a multilayer thin film is irradiated with white light, the light reflected from the sample is dispersed, and the reflection spectrum outside the measurement object is reduced based on the light source and the spectral pattern of each layer. A method for detecting the film thickness of a multilayer thin film, comprising: converting the spectrum into a wave number, and then performing a smoothing differentiation process to remove a DC component, and then performing a fast Fourier transform to detect the film thickness of the thin film. 2. A high-pass filter (y
_{2. The} method for detecting the thickness of a multilayer thin film according to claim 1, wherein the smoothing differentiation process is performed using _t ). (Equation 1) 3. The method for detecting a film thickness of a multilayer thin film according to claim 1, wherein a fast Fourier transform is performed by performing a process using a Hanning window (W _n ) represented by the following equation. (Equation 2) 4. The method for detecting a film thickness of a multilayer thin film according to claim 1, wherein a fast Fourier transform is performed by performing a process using a Slepian window which is an eigenvector of the matrix (A) represented by the following equation. (Equation 3)