JPH0815130A - Fourier transform phase modulation spectroellipsometry - Google Patents

Abstract

PURPOSE:To obtain a high speed high sensitivity infrared ellipsometry by combining Fourier transform infrared spectrophotometry with phase modulation spectroellipsometry. CONSTITUTION:A polarizer 3 and a phase modulation element 4 are arranged on the incident side of a sample 7 while an analyzer 5 and a light receiving unit 6 are arranged on the reflection side. A phase difference delta is introduced between Ppolarized light and S-polarized light from the polarizer 3 through modulation with frequency omega and the reflected light is detected through the analyzer 5. When two ellipsometric parameters at the time of reflection, i.e., the variation DELTA of phase difference between the P-polarized and S-polarized lights and the amplitude reflectance ratio angle PSI, are determined based on the DC component, and the components of frequencies omega, 2omega, an interferometer 2 for Fourier spectroscopy is arranged in the optical path between a light source 1 and time light receiving unit 6. The light receiving unit 6 detects the DC component and the components of frequency omega and 2omega from the light subjected to phase modulation and interference modulation. DELTA and PSI are determined based on the values obtained through Fourier transform of these three components.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to ellipsometry, and in particular, it is used for measuring ellipsometry parameters of a sample at high speed in a thin film forming process, etching process, surface cleaning process, etc. It relates to a possible Fourier transform spectroscopy phase modulation ellipsometry.

[0002]

2. Description of the Related Art When a substance is irradiated with light, the polarization state of incident light and the polarization state of reflected light are generally different as shown in FIG. This is because there is a difference between the reflectance and the phase shift at the time of reflection between P-polarized light (a polarized light component parallel to the incident surface) and S-polarized light (a polarized light component perpendicular to the incident surface). The change in the polarization state is represented by two parameters, that is, the change Δ in the phase difference between the P-polarized light and the S-polarized light and the amplitude reflectance ratio angle Ψ. Here, the complex reflection coefficient r _{P of} each of P-polarized light and S-polarized light,
The ratio of r _S r _P / r _S (complex reflectivity ratio), the two parameters delta, with Ψ is expressed as follows.

R _P / r _S = tan Ψ exp (iΔ) where Δ and Ψ are the wavelength of light, the angle of incidence, the complex refractive index of the substance,
Further, when there is a film, it is determined by the value of the film thickness or the like. Therefore, by measuring Δ and Ψ, the complex refractive index and the film thickness of the substance can be obtained.

The method for obtaining the above two parameters Δ and Ψ is ellipsometry, and the parameters Δ and Ψ are called ellipsometry parameters. Conventionally known polarization analysis methods include an extinction position detection method, a rotation analyzer method, and a phase modulation method. Among these, the extinction position detection method includes rotation of the polarizer and 1 /
This is a method of making incident elliptical polarized light in which the reflected light from the sample becomes linearly polarized light using a four-wave plate and rotating the analyzer on the reflection side so that the output of the detector becomes minimum and measuring Δ and Ψ. . In the rotating analyzer method, when linearly polarized light of 45 ° azimuth with respect to the incident surface is incident on the sample and the reflected light that is reflected and becomes elliptical polarized light is detected through an analyzer that rotates about the optical axis, the ellipse of the reflected light is obtained. Since a periodic signal reflecting the polarization state is obtained, this is a method of measuring Δ and Ψ. As shown schematically in FIG. 6, the phase modulation method uses a photoelastic modulator (Photo Elastic Modulator) on the incident side.
Or: PEM) is introduced, a phase difference δ modulated with a frequency ω is introduced between the P-polarized component and the S-polarized component of the light linearly polarized by the polarizer, and the reflected light is detected through an analyzer, This is a method of measuring Δ and Ψ from the DC component, the frequency ω component, and the frequency 2ω component.

In the infrared ellipsometry, the film thickness and the dielectric function (n-ik; n is the refractive index and k is the extinction coefficient) of the thin film can be determined at the same time, and at the same time, the infrared activity in the thin film is reduced. Various chemical bonds can be detected as resonance dispersion of the dielectric function. This analysis method can measure in situ changes in constituent chemical bonds during thin film formation, and it has been required to control the interface between the substrate and thin film and the surface state of the substrate and thin film at the atomic level. In recent semiconductor processes, it is playing an effective role. As an example of such an infrared ellipsometry, French Patent No. 86-11021 can be mentioned.

[0006]

The conventional infrared ellipsometry method is a combination of a Fourier transform infrared spectrometer and a rotating analyzer ellipsometer, or a dispersive infrared spectrometer and a phase modulation ellipsometer. It was done. The former is characterized by being able to disperse a wide range of wave numbers in a short time and with high sensitivity because it uses Fourier transform spectroscopy, but it performs phase modulation by a rotating analyzer and its rotation frequency is several tens of H.
There is a problem in that it is rate-determined to be about z and cannot follow a rapid change in the surface state.

The latter is capable of following a fast phenomenon because it performs phase modulation by a phase modulation element having a characteristic that a modulation frequency of about 50 kHz is possible. However, the phase-modulation ellipsometer is a method in which the target wave number range of the ellipsometry method using a dispersive spectrometer in the UV-visible region is changed to the infrared region, so it has low sensitivity and follows fast phenomena. Therefore, there is a problem that the wave number sweep width of the spectroscope must be narrowed.

When the Fourier transform spectroscope and the high speed phase modulation element are combined, the characteristics of both are utilized, and high speed and high sensitivity measurement can be realized. But in this case
Since the light during the movement of the movable mirror of the Michelson interferometer will be phase-modulated, the polarization analysis parameter must be obtained from the interferogram signal that has been phase-modulated. Such signal processing cannot be realized by using conventional signal processing methods such as Fourier transform infrared spectroscopy, rotary analyzer ellipsometry, and phase modulation ellipsometry.

The present invention has been made in view of the above problems of the prior art, and its object is to combine a Fourier transform infrared spectroscopy and a phase modulation ellipsometry to obtain a high-speed and high-sensitivity infrared polarization. It is an object of the present invention to provide a Fourier transform spectral phase modulation ellipsometry method that enables analysis, and in particular to provide a signal processing method for obtaining an ellipsometry parameter from an interferogram subjected to phase modulation.

[0010]

In the Fourier transform spectroscopic phase modulation ellipsometry of the present invention, the light emitted from the light source passes through a Michelson interferometer to become light in which all the wave numbers interfere, and a polarizer and a phase modulator After passing through, the light is reflected by the sample surface, and the reflected light is incident on the detector through the analyzer. The azimuth angles of the polarizer and the analyzer are set to ± 45 ° and ± 45 °, respectively. The azimuth angle of the phase modulation element is set to 0 ° or 90, and the phase difference between the polarized light in the azimuth direction and the polarized light perpendicular thereto is modulated at the frequency ω. In this method, unlike the conventional Fourier transform spectroscopy, among the light intensity signals output from the detector,
The DC component, the ω component, and the 2ω component are detected by the synchronous detection, and the interferometer movable mirror position dependency (interferogram) of these three signals is recorded. These interferograms are fast-Fourier-transformed and converted into the wave number dependence (spectrum) of DC component, ω component, and 2ω component. These three spectra can be converted into polarization parameter spectra by four arithmetic operations using the Bessel function value of the value of the modulation angle calibrated in advance.

When calibrating the modulation angle, the light emitted from the light source is
An optical system is constructed so that it is directly incident on the detector through the analyzer through the Michelson interferometer, the polarizer, and the phase modulation element. The azimuth angles of the polarizer and analyzer are set to ± 45 ° and ± 45 °, respectively, which are the same as those at the time of sample measurement, the phase modulator is set to 0 ° or 90, and the azimuth polarization and the polarization perpendicular thereto are set. The phase difference is modulated with the frequency ω. In the light intensity signal output from the detector, the DC component and the frequency 2ω component are detected by synchronous detection, and the interferogram of these two signals is recorded. These interferograms are fast-Fourier-transformed and converted into a DC component spectrum and a frequency 2ω component spectrum. The ratio of the spectral intensity of the DC component to the frequency 2ω component is a function of the modulation angle,
By solving the function for the modulation angle, the spectrum of the modulation angle is obtained, and the wavelength dependence of the polarization parameter can be calibrated.

That is, in the Fourier transform spectral phase modulation ellipsometry method of the present invention which achieves the above-mentioned object, a polarizer and a phase modulation element are sequentially arranged on the sample incident side, and an analyzer and a light receiver are sequentially arranged on the sample reflecting side. The phase difference δ modulated by the frequency ω is introduced between the P-polarized component and the S-polarized component of the light linearly polarized by the polarizer, and the reflected light is detected through the analyzer.
From the DC component, the frequency ω component, and the frequency 2ω component, two polarization analysis parameters for obtaining the two polarization analysis parameters of the change Δ in the phase difference between the P-polarized light and the S-polarized light at the time of reflection and the amplitude reflectance ratio angle Ψ are obtained. In the analysis method, an interferometer for Fourier spectroscopy is arranged in the optical path from the light source to the light receiver, and further,
The azimuth angle (P) of the polarizer is ± 45 °, the azimuth angle (M) of the phase modulation element is 0 ° or 90 °, and the azimuth angle of the analyzer is (A) ± 45 ° (these azimuth angles P , M, and A are arranged independently of each other), the light subjected to the phase modulation and the interferometric modulation is measured by the photodetector, and the direct-current component i _dc (x) of the obtained interfering optical signal, Frequency i component i
_I (3) obtained by detecting the three components of ₁ (x) and the component i ₂ (x) of frequency 2ω and performing Fourier transform of these three components
_This is a method characterized in that Δ and Ψ are obtained from _dc (k), I ₁ (k) and I ₂ (k) based on the following equations (33) and (34). ± sin 2Ψ (k) sin Δ (k) = [I ₁ (k) / 2J ₁ (δ ₀ (k))] / [I _dc (k) −I ₂ (k) J ₀ (δ ₀ (k)) ) / 2J ₂ (δ ₀ (k))] (33) ± sin 2Ψ (k) cos Δ (k) = [I ₂ (k) / 2J ₂ (δ ₀ (k))] / [I _dc (k) -I ₂ (k) J ₀ (δ ₀ (k)) / 2J ₂ (δ ₀ (k))] (34) where δ ₀ (k) is the phase modulation element It is the modulation amplitude, and the sign of ± in equations (33) and (34) is according to Table 2 below.

Further, with the sample removed, the analyzer and the light receiver are arranged in a transmission type in which the light transmitted through the phase modulator is directly incident, and the azimuth angles of the polarizer, the phase modulator and the analyzer are set. With the same condition, the light that has undergone phase modulation and interferometric modulation is measured by the photodetector, and the direct current component of the obtained interfering optical signal,
It is possible to detect the two components of the frequency 2ω, obtain the wave number dependence of the modulation amplitude of the phase modulation element from the values obtained by Fourier transforming these two components, and use it to obtain Δ and Ψ.

[0014]

In the present invention, in the phase modulation ellipsometry, an interferometer for Fourier spectroscopy is arranged in the optical path from the light source to the light receiver, and the light subjected to the phase modulation and the interference modulation is measured by the light receiver. Then, three components of the obtained interference light signal, that is, the DC component, the frequency ω component, and the frequency 2ω component are detected, and the P polarization and S polarization at the time of reflection are calculated from the values obtained by Fourier transforming these three components. Phase difference change Δ and amplitude reflectance ratio angle Ψ
Since the two polarization analysis parameters are obtained, it is possible to measure the polarization analysis parameters in a wide range of wave numbers at high speed and with high sensitivity.

In addition, when the measurement of the sample shifts to the measurement for the wave number dependence of the modulation amplitude of the phase modulation element, or vice versa, the azimuths of the polarizer, the phase modulation element, and the analyzer are respectively changed. Since the angle does not have to be changed at all, there are few fluctuations in the arrangement, and accurate and short-time measurement is possible.

[0016]

DESCRIPTION OF THE PREFERRED EMBODIMENTS First, a conventional dispersion type phase modulation method as shown in FIG. 6 will be described. Reflected light for a certain wave number k is given by (Rev. Sci. Instru
m. 53 (7), Jul. 1982, pp. 969-9
77).

I (k, t) = I ₀ (k) + I _S (k) sin [δ (k, t)] + I _C (k) cos [δ (k, t)] (1) where Then, the modulation angle (phase difference) δ (k, t) and each coefficient in the PEM are as follows: δ (k, t) = δ ₀ (k) sin ωt (2) I ₀ (k) = K ( k) [1 + cos 2Acos 2Mcos 2 (M-P) - (cos 2A + cos 2Mcos 2 (M-P)) cos 2Ψ- sin 2Asin 2Mcos 2 (M-P) sin 2sin 2Ψcos Δ] ··· (3) I S ( k) = K (k) [ - sin 2 (M-P) sin 2Asin 2Ψsin Δ] ··· (4) I C (k) = K (k) [- sin 2 (M-P) {sin 2M ( cos 2Ψ− cos 2A) + cos 2Msin 2Asin 2Ψsin Δ}] (5) K (k) = | r _P (k) ² + r _S (k) ² | / 4 (6) Here, azimuth angles P, P of the pass axis of the polarizer
Azimuth M of pressure axis of EM, azimuth A of pass axis of analyzer
Is measured from the incident surface when the xz plane is taken as the incident surface as shown in FIG.

As a special case, P = ± 45 °, M = 0
Or 90 °, A = ± 45 ° (P, M, and A are mutually independent), I ₀ (k) = K (k) (7) I _S (k) = ± K (k ) [Sin 2Ψsin Δ] (8) I _C (k) = ± K (k) [sin 2Ψcos Δ] (9) However, the signs of ± in formulas (8) and (9) are as shown in Table 1 below.

[0019] Therefore, by determining I ₀ (k), I _S (k), and I _C (k), Ψ (k) and Δ (k) can be determined. When the reflected light is limited to the frequency range of 2ω in the general range, the following is obtained.

I (k, t) = I _dc (k) + I ₁ (k) sin ωt + I ₂ (k) cos 2ωt (10) Here, each coefficient is I _dc (k) = I ₀ ( k) + I _C (k) J ₀ (δ ₀ (k)) ・ ・ ・ (11) I ₁ (k) = 2I _S (k) J ₁ (δ ₀ (k)) ・ ・ ・ (12) I ₂ (K) = 2I _C (k) J ₂ (δ ₀ (k)) (13) However, J ₀ , J ₁ , and J ₂ are each 0
It is a Bessel function of the first order, second order, and second order.

Therefore, by determining the DC component I _dc (k), ω component I ₁ (k), and 2ω component I ₂ (k) by measurement, I ₀ (k), I _S (k), and I _C (K) is obtained as follows.

I ₀ (k) = I _dc (k) −I ₂ (k) J ₀ (δ ₀ (k)) / 2J ₂ (δ ₀ (k)) (14) I _S (k) = I ₁ (k) / 2J ₁ (δ ₀ (k)) ・ ・ ・ (15) I _C (k) = I ₂ (k) / 2J ₂ (δ ₀ (k)) ・ ・ ・ (16) Values of equations (14) to (16) and equations (7) to (9)
From the relationship of, Ψ (k) and Δ (k) can be obtained. Of course, the formulas (7) to (9) are in a special case, but not limited thereto, and general formulas (3) to (5).
Based on the above, Ψ (k) and Δ (k) can be obtained.
Rev. described above. Sci. Instrum. 53
(7), Jul. 1982, pp. In 969-977, δ so that J ₀ (δ ₀ (k)) = 0.
Setting ₀ (k) to 2.405 rad, the above equation (1
4) to (16) are simplified.

The above method cannot be used as it is in the case of Fourier transform spectroscopy which takes in light of all wave numbers at once. As one of the proposals, the following method can be considered. The reflected light intensity before the Fourier transform is expressed by the formula (1
Inverse Fourier transform of 0) is as follows.

I (x, t) = i _dc (x) + i ₁ (x) sin ωt + i ₂ (x) cos 2ωt (17) Here, each coefficient is the inverse of each coefficient of the equation (10). Fourier transformed, i (x, t) = F ⁻¹ [I (k, t)] (18) i _dc (x) = F ⁻¹ [I _dc (k)] ( 19) i ₁ (x) = F ⁻¹ [I ₁ (k)] (20) i ₂ (x) = F ⁻¹ [I ₂ (k)] (21)

Therefore, in the case of Fourier transform spectroscopy, I
To obtain _dc (k), I ₁ (k), and I ₂ (k), the DC component of the interferogram signal before Fourier transform, ω
It suffices to measure the component and the 2ω component and perform Fourier transform on each separately. In order to calculate the values of Ψ (k) and Δ (k), I _dc (k), I ₁ (k), and I ₂ (k)
From equations (14) to (16), I ₀ (k), I
_It is necessary to calculate _S (k) and I _C (k). that time,
It is necessary to use δ ₀ (k), that is, the PEM modulation amplitude. However, it is general that the modulation angle of the PEM has wave number dependence, and as in the above-mentioned prior art, δ
_Since it is impossible in principle to fix ₀ (k) = 2.405 rad in the case of Fourier transform spectroscopy, δ
_It must be assumed that ₀ (k) is different for each wave number.

However, if δ ₀ (k) is known for each wave number, as shown in the flowchart of FIG. 4, in step ST1, the band of the interferogram signal obtained from the detector is limited to 2ω, Obtain the signal of (17),
Next, in step ST2, the DC component i _dc (x), ω component i ₁ (x), and 2ω component i ₂ (x) of the signal thus obtained are separated, and in step ST3, the respective components are separated. Fourier transform is performed separately to obtain I _dc (k), I ₁ (k),
I ₂ (k) is obtained, and then at step ST4, a known δ
_With reference to the table of ₀ (k), equations (14) to (1
6) is used to calculate I ₀ (k), I _S (k), and I _C (k), and in step ST5, the polarization analysis parameters Ψ (k), Δ are calculated using the equations (7) to (9). (K) can be obtained (in FIG. 4, P = −45 °, M = 0 °, A =
We assume -45 °. ). Note that in a general arrangement, instead of formulas (7) to (9), general formulas (3) to
Using (5), Ψ (k) and Δ (k) can be obtained.

From the above, it can be said that the Fourier transform spectral phase modulation ellipsometry can be realized if the modulation amplitude δ ₀ (k) of the PEM is known in advance. Therefore, in the present invention, the sample is arranged so that the incident light directly reaches the detector, and this δ is set as follows.
₀ (k) is obtained, and using this, the polarization analysis parameters Ψ (k) and Δ (k) are obtained in the order shown in FIG.

The above equations (1) to (16) are equations for the polarization analysis of the reflection method. The change in the polarization state received by the incident light before reaching the detector is represented by Ψ and Δ. Then, the case of the transmissive method also holds. In the transmission method, if nothing is placed at the position where the sample should be placed, there is no change in the polarization state, so r _P (k) = 1 ... (22) r _S (k) = 1 ... (23) tan Ψexp (iΔ) = r _P (k) / r _S (k)
Therefore, tan Ψ = 1, that is, Ψ = 45 ° (24) Δ = 0 ° (25). At this time, as the optical system, P = ± 45 °, M = 0 ° or 90 °, which is the same as the measurement by the reflection method in which the sample is arranged,
Assuming that the arrangement is A = ± 45 ° (P, M, and A are mutually independent), I ₀ (k) = K (k) (26) I _S (k) = ± K (k) [sin 2Ψsin Δ] = 0 (27) I _C (k) = ± K (k) [sin 2Ψcos Δ] = ± K (k) (28) However, the sign of ± in Expression (28) is determined by the same relationship as Expression (9) in Table 1. Therefore, the formula (1
0), the DC, ω, and 2ω components of the output signal of the detector are expressed as follows: I _dc (k) = K (k) (1 ± J ₀ (δ ₀ (k ))) (29) I ₁ (k) = 0 (30) I ₂ (k) = ± 2K (k) J ₂ (δ ₀ (k)) (31) .

Therefore, from the equations (29) and (31), I ₂ (k) / I _dc (k) = ± 2J ₂ (δ ₀ (k)) / (1 ± J ₀ (δ ₀ (k)) ) ... (32) If this equation (32) is solved for δ ₀ (k),
Δ from the DC component and 2ω component of the signal from the detector
₀ (k) is obtained.

Even when Fourier transform spectroscopy is performed, I _dc (k) and I ₂ (k) can be obtained from the Fourier transform values of i _dc (x) and i ₂ (x) of the interferogram.
Δ ₀ (k) can be obtained by using the equation (32).

Of course, equations (26) to (28) are for special arrangements, and in general, equations (3) to (5) are replaced by equation (2).
4) and (25), by substituting Ψ = 45 ° and Δ = 0 °, I ₀ (k), I _S (k), and I _C (k) are obtained, and their values and the equation (11) are obtained. ~ (13) gives I _dc (k),
I ₁ (k) and I ₂ (k) are obtained as a function of δ ₀ (k), and they are simultaneous to obtain δ ₀ (k).

From the above, δ ₀ (k) can be obtained from the DC component and 2ω component of the signal from the detector by performing measurement in the optical system in which the polarizer and the analyzer face each other. Even if there is a window in the middle, if the window is perpendicular to the optical axis, the effect of the window on P-polarized light and S-polarized light is the same, so the polarization state does not change.

Among the above special arrangements, P = -45 °, M
= 0 ° and A = −45 °.
I ₂ (k) / I _{dc with} respect to the modulation amplitude δ ₀ (k) of EM)
The change in (k) is shown in FIG. From this figure, it can be seen that if one I ₂ / I _dc is determined, the corresponding δ _{0 is} uniquely determined.

As described above, except for the sample arrangement,
Arrange the incident light so that it reaches the detector directly.
₀ (k) is obtained, and using this, the polarization analysis parameters Ψ (k) and Δ (k) can be obtained according to the flowchart of FIG.

An embodiment of the Fourier transform spectral phase modulation ellipsometry method of the present invention will be described below with reference to FIGS. 1 and 2. FIG. 1 shows the configuration of an apparatus for carrying out ellipsometry according to an embodiment of the present invention. The illustrated apparatus constitutes a Fourier transform infrared spectroscopic phase modulation polarization analyzer in which the phase modulation element 4 is inserted in a Fourier transform infrared spectroscope, and includes a light source 1 and infrared light from the light source 1 including a movable mirror. Michelson interferometer 2 for receiving an interferogram and a polarizer 3 having an azimuth angle (P) of ± 45 ° provided at a position where the modulated light emitted from the Michelson interferometer 2 is incident, (M) From the surface of the sample 7 arranged at the position where the phase modulation element (PEM) 4 having 0 ° or 90 ° and modulating the phase difference at the frequency ω, and the light phase-modulated by the phase modulation element 4 are incident. An analyzer 5 having an azimuth angle (A) of ± 45 ° in the reflected optical path, and a light receiver 6 for photoelectrically converting the light passing through the analyzer 5.
And 8 and 8 ′ collect light on the surface of the sample 7,
Further, concave mirrors that convert the light diverging from the irradiation point on the surface of the sample 7 into parallel light, and 9 and 9 ′ are plane mirrors that bend the optical path.

In the apparatus of FIG. 1, the time t dependence of the light intensity signal i measured at a certain moving mirror position x is given by the following equation. i (x, t) = i _dc (x) + i ₁ (x) sin ωt + i ₂ (x) cos 2ωt (17) where i _dc (x), i ₁ (x), i ₂ (x ) Are the DC component, the ω component, and the 2ω component of the signal obtained by the light receiver 6, respectively. When these three components are synchronously detected and independently Fourier-transformed, the wave number k dependence I of the three components I
It is converted into _dc (k), I ₁ (k), and I ₂ (k). From these equations (14) to (16) and equations (7) to (9), the following relationships exist between these values and the polarization analysis parameters Ψ (k) and Δ (k).

± sin 2Ψ (k) sin Δ (k) = [I ₁ (k) / 2J ₁ (δ ₀ (k))] / [I _dc (k) −I ₂ (k) J ₀ (δ ₀ (K)) / 2J ₂ (δ ₀ (k))] (33) ± sin 2Ψ (k) cos Δ (k) = [I ₂ (k) / 2J ₂ (δ ₀ (k))] / [I _dc (k) -I ₂ (k) J ₀ (δ ₀ (k)) / 2J ₂ (δ ₀ (k))] (34) However, in equations (33) and (34) The sign of ± is as shown in Table 2 below.

[0038] Therefore, from the above equations (33) and (34), Ψ (k), Δ
The spectrum of (k) is obtained, where δ ₀ (k)
Is the wave number dependence of the modulation amplitude of the modulation angle in the phase modulation element 4. J ₀ (δ ₀ (k)), J ₁ (δ ₀ (k)), J ₂
(Δ ₀ (k)) are Bessel functions of 0th order, 1st order, and 2nd order, respectively.

The spectrum of Ψ (k) and Δ (k) is calculated by the equation (3)
3), δ ₀ (k) required to obtain by (34)
Is determined as follows. FIG. 2 shows an apparatus for carrying out the first embodiment. This device constitutes a Fourier transform infrared spectroscopic phase modulation polarization analyzer in which the phase modulation element 4 of FIG.
, A Michelson interferometer 2 including a movable mirror for receiving infrared light from the light source 1 to form an interferogram, and an azimuth angle (P) provided at a position where the modulated light emitted from the Michelson interferometer 2 is incident. ) A polarizer 3 having ± 45 °, a phase modulation element (PEM) 4 having an azimuth angle (M) of 0 ° or 90 ° and modulating a phase difference with a frequency ω, and a phase modulation element 4 subjected to phase modulation A sample 5 arranged at a position where light is incident, an analyzer 5 having an azimuth angle (A) of ± 45 ° in an optical path reflected from the surface of the sample, and a photodetector 6 for photoelectrically converting light passing through the analyzer 5.
And have. This apparatus uses the same light source 1, Michelson interferometer 2, polarizer 3, analyzer 5, and light receiver 6 shown in FIG. 1, except for the sample 7 of the reflection type shown in FIG. 3, the azimuth angles of the phase modulation element 4 and the analyzer 5 are not changed and the arrangement is changed to the transmission method.

Also in the apparatus shown in FIG. 2, the time t dependency of the light intensity signal i measured at a certain moving mirror position x is given by the equation (17). In the same manner as in the case of FIG. 1, among the three components subjected to the synchronous detection and the Fourier transform and having the wave number k, the ratio of I _dc (k) and I ₂ (k) is given by the following equation. .

I ₂ (k) / I _dc (k) = ± 2J ₂ (δ ₀ (k)) / (1 ± J ₀ (δ ₀ (k))) (32) Therefore, from the above equation, δ ₀ (k) is obtained. Further, from the obtained δ ₀ (k), J ₀ (δ ₀ (k), J
_{Since 1} (δ ₀ (k)) and J ₂ (δ ₀ (k)) can be calculated in advance, Ψ (k) and Δ (k) can be calculated using equations (33) and (34). At the time of calculation, it is not necessary to calculate the Bessel function from δ ₀ (k).

In the above, the polarizer 3, the analyzer 5 and the phase modulation element 4 are set at a specific azimuth angle. However, as described above, the reflection type arrangement including the sample 7 and the phase modulation element 4 are set. In the transmission type arrangement for obtaining the wave number dependence δ ₀ (k) of the modulation amplitude of, even if these azimuth angles are set to other azimuth angles, equations (3) to (5) and equation (11) to (1
Ψ (k) and Δ (k) can be obtained from 3), and I ₀ (k) and I can be obtained by substituting Ψ = 45 ° and Δ = 0 ° into equations (3) to (5). _S (k), I _C (k) is obtained, and I _dc (k),
Although I ₁ (k) and I ₂ (k) can be obtained as a function of δ ₀ (k) and they can be simultaneously obtained to obtain δ ₀ (k), in these cases, the polarizer 3, the phase modulation It is necessary to change the azimuth angle of each of the element 4 and the analyzer 5 when changing from the reflection type arrangement to the transmission type arrangement, the readjustment takes time and is complicated, and the measurement is interrupted for a long time. This is not preferable. On the other hand, in the method of the present invention in which these azimuth angles are not changed at the time of both measurements, there are few fluctuation points and accurate measurement is possible.
₀ (k) can be obtained, which is a preferable method.

The arrangement position of the Michelson interferometer 2 is not limited to between the light source 1 and the polarizer 3, and the light source 1
It may be provided at any position in the optical path from the light receiver 6 to the light receiver 6.

In this measuring device, since the reflecting mirrors 9, 8, 8 ', 9'are inserted in the measuring optical path, the complex reflectance ratio of the entire portion sandwiched between the phase modulation element 4 and the analyzer 5 is ta
n ψ _eff exp (iΔ _eff ) is measured. If the complex reflectance ratio of the sample 7 is tan Ψ exp (iΔ) and the total complex reflectance ratio of all reflecting mirrors is tan Ψ _M exp (iΔ _M ), tan Ψ _eff exp (iΔ _eff ) = tan Ψ _M exp (IΔ
_M ) tan Ψ exp (iΔ). Therefore, the complex reflectance ratio of only the sample 7 is measured by using the sample 7 of which the complex reflectance ratio is known, and tan Ψ _M exp (iΔ _M ) = {tan Ψ _eff exp (iΔ _eff )} / { tan Ψ exp (iΔ)} (35) In addition, since the optical density that is usually measured by infrared ellipsometry is calculated using the ratio of the case where there is no thin film on the substrate and the case where there is a thin film on the substrate, the complex reflectance of the reflecting mirror regardless of the presence or absence of a thin film. The effect of is eliminated by taking the ratio and does not require calibration.

In the above first embodiment, δ
_{When 0} (k) was determined, the same light source 1, Michelson interferometer 2, polarizer 3, phase modulation element 4, analyzer 5 and light receiver 6 as those used for the sample measurement were used. The same light source 1, Michelson interferometer 2, polarizer 3, and phase modulation element 4 are used for sample measurement, and another analyzer 5'and a light receiver 6'are used as an analyzer and a light receiver. You may use. It is also important to be able to adjust the incident angle of the measurement light on the sample 7. The embodiment of FIG. 3 is an embodiment having both of them, and the same components as those of FIG. 1 are denoted by the same reference numerals. In this example, Example 1
Similarly to, the light source 1, the Michelson interferometer 2 that includes the moving mirror and receives the infrared light from the light source 1 to form an interferogram, and the modulated light emitted from the Michelson interferometer 2 are provided at positions to be incident. Polarizer with azimuth angle (P) ± 45 ° 3
A phase modulator (PEM) 4 having an azimuth angle (M) of 0 ° or 90 ° and modulating a phase difference with a frequency ω; and a plane mirror 9 for bending an optical path of light phase-modulated by the phase modulator 4.
Concave mirror 8 that collects reflected light from plane mirror 9 on the surface of sample 7
A sample 7 arranged at a condensing position, a concave mirror 8'for converting the light reflected from the surface of the sample 7 into parallel light, and a concave mirror 8 '.
Plane mirror 9'for bending the optical path of parallel light from
Azimuth angle (A) ± 45 ° arranged in the optical path reflected from
And an optical receiver 6 that photoelectrically converts light that has passed through the analyzer 5.

In this embodiment, the plane mirror 9 on the incident side of the sample 7 is switchable between the position indicated by the solid line and the position indicated by the broken line. The intensity signal i of equation (17) is obtained in the same manner as in 1, and the DC component, ω component, and 2ω component thereof are Fourier transformed to obtain I _dc (k), I ₁ (k), and I ₂ (k). , (33) and (34), the spectra of Ψ (k) and Δ (k) are obtained. Further, when the plane mirror 9 is switched to the position of the broken line, the light that has been phase-modulated by the phase modulation element 4 enters the analyzer 5 ′ and the light receiver 6 ′ of another set so that the analyzer 5 ′ and the light receiver 6 ′. 6'is arranged. The azimuth angle (A) of the analyzer 5 ′ in this case is ± 45 ° and is fixed. Using this analyzer 5 ′ and the light receiver 6 ′, δ ₀ (k) can be obtained from the equation (32) in the same manner as in the case of FIG. In this case, the azimuth angles of the polarizer 3, the phase modulation element 4, the analyzer 5 and the analyzer 5 ′ are all fixed at the time of both measurements, and only the plane mirror 9 changes the azimuth angle. Will be possible.

In the case of FIG. 3, in order to measure the incident angle φ dependence of the polarization analysis parameters Δ and Ψ, the concave mirrors 8 and 8 ′ are respectively plane mirrors 9 as shown by the one-dot chain line in the figure. Can be moved by the same distance along the reflected light path and the incident light path to the plane mirror 9 ', and accordingly, the sample 7 also moves upward in the figure. The concave mirrors 8 and 8'are rotatable in opposite directions so that the reflected light path from the concave mirror 8 and the incident light path to the concave mirror 8'cross each other on the surface of the sample 7. φ
Can be changed continuously. The movement and rotation of the concave mirrors 8 and 8'and the sample 7 as described above are performed, for example, in Japanese Patent Publication No. 47-3991.
It can be easily realized by using a link mechanism based on the publication No. 1. As a mechanism for continuously changing the incident angle φ on the sample 7 as described above, Japanese Patent Publication No. 47-39911 is available.
The present invention is not limited to that disclosed in Japanese Patent Publication No. 2-48848, and various known mechanisms can be used.

The Fourier transform spectral phase modulation ellipsometry method of the present invention has been described above based on its principle and embodiments, but the present invention is not limited to these embodiments and various modifications can be made.

[0049]

As is apparent from the above description, according to the Fourier transform spectroscopy phase modulation ellipsometry of the present invention, in the phase modulation ellipsometry, the Fourier transform spectroscopy for Fourier spectroscopy is performed in the optical path from the light source to the light receiver. An interferometer is arranged, and the light subjected to the phase modulation and the interference modulation is measured by the photodetector, and the three components of the obtained interference light signal, that is, the DC component, the frequency ω component, and the frequency 2ω component are detected, and these 3 components are detected. Since two polarization analysis parameters of the phase difference Δ between the P-polarized light and the S-polarized light of the reflected light and the reflection amplitude ratio angle Ψ are obtained from the value obtained by Fourier transforming the component,
It becomes possible to measure ellipsometry parameters in a wide range of wave numbers at high speed and with high sensitivity.

In addition, when the measurement of the sample shifts to the measurement for the wave number dependence of the modulation amplitude of the phase modulation element, or vice versa, the azimuths of the polarizer, the phase modulation element, and the analyzer are respectively changed. Since the angle does not have to be changed at all, there are few fluctuations in the arrangement, and accurate and short-time measurement is possible.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration of an apparatus for carrying out ellipsometry according to an embodiment of the present invention.

FIG. 2 is a diagram showing an arrangement for obtaining a wave number dependence of a modulation amplitude of a modulation angle.

FIG. 3 is a diagram showing a configuration of an apparatus for performing ellipsometry according to another embodiment of the present invention.

FIG. 4 is a flowchart of a calculation for obtaining a polarization analysis parameter by the polarization analysis method of the present invention.

FIG. 5 is a graph for obtaining a modulation amplitude of a modulation angle.

FIG. 6 is a diagram showing an arrangement for a conventional distributed phase modulation method.

FIG. 7 is a diagram for defining azimuth angles of a polarizer, a phase modulation element, and an analyzer.

FIG. 8 is a diagram schematically showing a polarization state of incident light and a polarization state of reflected light when a substance is irradiated with light.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Light source 2 ... Michelson interferometer 3 ... Polarizer 4 ... Phase modulation element (PEM) 5, 5 '... Analyzer 6, 6' ... Photoreceiver 7 ... Sample 8, 8 '... Concave mirror 9, 9' ... Plane mirror

Claims (2)

[Claims] 1. A polarizer and a phase modulator are sequentially arranged on the sample entrance side, and an analyzer and a photodetector are sequentially arranged on the sample reflection side, and a P-polarized component of light linearly polarized by the polarizer and S A phase difference δ modulated with frequency ω is introduced between polarization components, reflected light is detected through an analyzer, and its DC component and frequency ω
P component and S component at the frequency of 2ω
In a phase-modulation ellipsometry method for obtaining two polarization analysis parameters of a change Δ in the phase difference of polarization and an amplitude reflectance ratio angle Ψ, an interferometer for Fourier spectroscopy is arranged in the optical path from the light source to the light receiver. And further, the azimuth angle (P) of the polarizer
Is ± 45 °, the azimuth angle (M) of the phase modulator is 0 ° or 90 °, and the azimuth angle of the analyzer is (A) ± 45 ° (these azimuth angles P, M, and A are mutually independent). The light having undergone the phase modulation and the interferometric modulation is measured by the photodetector in the state of being arranged so that the DC component i of the obtained interfering optical signal is
_dc (x), component i ₁ frequency ω (x), to detect the three components of component i ₂ (x) of the frequency 2 [omega, obtained by these three components by Fourier transform _{I dc (k), I 1} ( k), I
₂ Fourier transform spectroscopic phase modulation ellipsometry characterized by obtaining Δ and Ψ from (k) based on the following equations (33) and (34). ± sin 2Ψ (k) sin Δ (k) = [I ₁ (k) / 2J ₁ (δ ₀ (k))] / [I _dc (k) −I ₂ (k) J ₀ (δ ₀ (k)) ) / 2J ₂ (δ ₀ (k))] (33) ± sin 2Ψ (k) cos Δ (k) = [I ₂ (k) / 2J ₂ (δ ₀ (k))] / [I _dc (k) -I ₂ (k) J ₀ (δ ₀ (k)) / 2J ₂ (δ ₀ (k))] (34) where δ ₀ (k) is the phase modulation element It is the modulation amplitude, and the sign of ± in equations (33) and (34) is according to Table 2 below. 2. The polarizer according to claim 1, wherein the analyzer and the light receiver are arranged in a transmission type in which light transmitted through the phase modulation element is directly incident with the sample removed.
With the azimuth angles of the phase modulation element and the analyzer kept the same, the light subjected to the phase modulation and the interferometric modulation is measured by the photodetector, and the direct current component and the frequency 2ω component of the obtained interference optical signal are measured. Fourier transform spectroscopy characterized by detecting two components, obtaining the wave number dependence of the modulation amplitude of the phase modulation element from the values obtained by Fourier transforming these two components, and using this to obtain Δ and Ψ. Phase modulation ellipsometry.