JP2000275016A - Measurement method and measuring apparatus for film characteristic value distribution - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a measurement method and a measuring apparatus for a film characteristic value distribution whereby a thickness distribution of a thin film sample can be linearly obtained without scanning a light. SOLUTION: A convergent slit light passing through a cylindrical lens 17 is irradiated to a sample 11 and a reflecting light from the sample 11 enters a, CCD camera 20. The CCD camera 20 consists of a two-dimensional image sensor having one axis agreed with an axis direction of the convergent slit light and the other axis agreed with a direction of an angle of incidence. Information from the CCD camera 20 is input to a computer 21. The computer 21 fits a reflectance R related to a thickness (d) of the sample 11 and an angle of incidence (q) with a reflectance R1 related to a thickness (d1) and an angle of incidence (q1) and, obtains the thickness (d) of the sample 11.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for measuring a film characteristic value distribution for obtaining a film thickness of a thin film sample along at least a straight line.

[0002]

2. Description of the Related Art The thickness of a thin film is an important parameter for knowing its optical properties.
With the development of thin film technology in the fields of optoelectronics and optics, a number of film thickness measurement methods have been proposed. Variable angle reflection (VAR), which measures the incident angle dependence R (θ) of the reflectivity of the thin film sample and analyzes the result using an algorithm based on the Fresnel equation to obtain the film thickness
Has been applied in various forms to meet the needs of the site at that time.

For recently demanded "high-speed" measurement, a beam profile reflectometry (BPR) has been proposed in which the VAR eliminates a mechanical drive for scanning the incident angle. This is a method in which convergent light is incident on a thin film and reflected light is received by a one-dimensional image sensor to measure R (θ) at a time.

[0004]

However, recently, there has been a demand not only for speeding up the measurement but also for measuring the film thickness distribution in order to examine the film thickness unevenness in the film coating such as the antireflection film of the optical element. Is growing. In contrast, BP
The general film thickness measurement method including R employs a point measurement method. Therefore, in order to obtain the film thickness by distribution, position scanning of the sample or the measurement pattern is required, but this leads to an increase in measurement time and a decrease in measurement accuracy.

The present invention has been made in view of the above points, and is intended to measure a film characteristic value distribution capable of obtaining at least a film thickness along a straight line of a thin film sample without performing mechanical scanning. It is an object to provide a method and a measuring device.

[0006]

According to the present invention, a thin film sample is irradiated with convergent slit light passing through a cylindrical lens, and light reflected from the sample is made to coincide with one axis in the major axis direction of the convergent slit light. A step of detecting with a two-dimensional image sensor that is made to coincide with the other incident angle dependent characteristic detection direction, based on information from the two-dimensional image sensor, the position on the slit (y), and the thickness of the sample (d ) And the angle of incidence on the sample
(θ) and the reflectance R (y, θ), the thickness (d) of the sample, the refractive index (n), and the theoretical reflectance (RT (θ:
n, d) and a step of determining at least the thickness d (y) along the axial direction of the convergent slit light of the sample, and a method for measuring a film characteristic value distribution, and a thin film sample. A thin-film sample holding unit to be held, a light source, a cylindrical lens that converges light from the light source and projects convergent light onto the thin-film sample on the thin-film sample holding unit, and receives reflected light from the sample, A two-dimensional image sensor in which one axis is made to coincide with the long axis direction of the convergent light and the other axis is made coincident with the incident angle dependent characteristic detection direction. Position y, the reflectance R (y, θ) according to the incident angle (θ) to the sample, and the thickness of the sample (d)
And the refractive index (n) of the sample and the theoretical reflectance RT (θ: n, d) related to the incident angle (θ) by fitting at least the thickness of the sample along the major axis direction of the convergent slit light. and a calculation unit for calculating d (y).

According to the present invention, the thickness of a thin film sample can be easily obtained along a straight line.

[0008]

Embodiments of the present invention will be described below with reference to the drawings. 1 to 7 are views showing a method and apparatus for measuring a film characteristic value distribution according to the present invention.

As shown in FIG. 1, a film characteristic value measuring device 10 includes a thin film sample holder 12 for holding a thin film sample 11.
, A light source 13, a polarization direction switching unit (polarizer) 14, a spatial filter 15, a lens 16, and a cylinder that converges light from the light source 13 and emits convergent slit light to the thin film sample 11. And a lens 17.

The reflected light from the sample 11 is
2D image sensor 2 consisting of CCD camera after 19
0 is detected. The two-dimensional image sensor 20 has one axis oriented in the major axis direction y of the convergent slit light to the sample 11.
(FIG. 1) and the other axis coincides with the incident angle dependent characteristic detection direction θ (FIG. 2). Here, the incident angle dependent characteristic detection direction θ is a normal to the sample 11.
This is the angle between N and the reflected light.

The two-dimensional image sensor 20 is connected to an arithmetic unit (computer) 21 for calculating the thickness and the refractive index of the sample 11 based on information from the two-dimensional image sensor 20.

Next, the operation of the embodiment having the above-described configuration will be described. In FIG. 1, a He-Ne laser having a wavelength of 632.8 nm is used as the light source 13. The light emitted from the light source 13 is converted into linearly polarized light by using the polarizing element 14, and the spatial filter 15 and the lens 16 (f = 25
0, f = 30 mm) to increase the beam diameter. Here, the polarizing element 14 is composed of a polarizer or a wave plate that switches the polarization direction between p-polarized light and s-polarized light temporally.

The light passes through a cylindrical lens 17 (f = 50 mm, 30 × 3).
0 mm ² ) and the sample 11 has an angular width θw = ± 14 m, for example.
Incident. On sample 11, 0.75 mm width and length (y
Direction) 30 mm convergent slit light is obtained. The reflected light from the sample 11 is reflected by the cylindrical lens 18 (f = 50 mm, 30 × 30 m
m2), the light is converted into a parallel light, and a pair of cylindrical lenses 19 (f = 50 m
m, 30 × 30 mm 2), the reflectance distribution image captured by the two-dimensional CCD camera 20 is taken into the computer 21 as image data of 480 pixels in the y direction × 512 pixels in the θ direction.
With such an optical system, the central incident angle of the incident convergent light with respect to the sample 11 is set to 50 °, and the incident angle (θ) is changed from 36 ° to 64 °.
Measure the reflectance curve for each sample position y against ° without mechanical scanning. The measurement time is determined by the time required for reading one image, and is 1/30 second.

The obtained reflectance distribution image is stored in a computer 21.
Is performed to obtain the reflectance angle dependency and the film thickness of each point on the slit image. The time required for the analysis is about 4 minutes for the data of 480 pixels in the y direction and 441 pixels in the θ direction used in the experiment.

The analysis for obtaining the film thickness and the refractive index in the computer 21 is performed as follows. Ie CCD
Based on the information from the camera 20, the reflectance R (y, θ) related to the incident angle (θ) to the sample at each point y of the sample 11 is obtained, and at the same time, the thickness (d) of the sample 11 and the refractive index ( n.) and the theoretical reflectance RT (θ: n, d) relating to the incident angle (θ). Next, by fitting R and RT using the following equation, the thickness d (y) and the refractive index n (y) of the sample 11 along the major axis direction y of the convergent slit light can be obtained. .

[0016]

(Equation 1) However, is a theoretical expression of the reflectance of the single-layer film, and the following is given with respect to the reflection coefficient ρ and the transmission coefficient τ at the two interfaces and the phase shift 2ψ due to the reciprocation of the film (FIG. 2).

[0017]

(Equation 2) Here, the variables on the right side are given as follows, where the refractive index of air is 1, the refractive index of the substrate is n0, and the refractive angles of the film and the substrate are φ and φ0, respectively. First, the law of refraction is established between the incident angle and the refraction angle.

[0018]

(Equation 3) The phase shift due to film reciprocation is

[0019]

(Equation 4) And the reflection coefficient is

[0020]

(Equation 5) For s-polarized light

[0021]

(Equation 6) The transmission coefficient for p-polarized light is

[0022]

(Equation 7) For s-polarized light

[0023]

(Equation 8) In the above, when the refractive index n of the sample 11 is known in advance, the refractive index (n) of the sample 11 is not determined.
May be determined only for the thickness (d).

The polarizer 14 is used to switch the polarization direction of the convergent light incident on the sample 11 from the cylindrical lens 17 to s-polarized light and p-polarized light. And R and RT for P-polarized light
Thereby, fitting accuracy can be further improved.

Next, the calibration of the measurement data in the computer 21 will be described. When the reflectance distribution image is captured by the CCD camera 20, the long axis direction of the convergent slit light
Since the information on the position of y and the information on the incident angle θ are represented by pixel numbers (P, Q), it is necessary to convert them to θ and y.
In the slit axis direction y, since the cylindrical lenses 18 and 19 do not have an image forming action, the number of pixels in the y direction (480 pixels) and the length of one pixel (13 mm) indicate a straight line corresponding to 6.24 mm. It can be seen that the film thickness distribution can be measured, and the absolute relationship with Q can be determined from the position where the sample is fixed.

The angle of incidence θ is directly related to the accuracy and result of obtaining the film thickness by fitting the obtained reflectance curve to a theoretical formula, and therefore the absolute relationship with P is accurately obtained. There is a need. The cover glass (thickness d = 149 mm, refractive index)
Using n = 1.52), a conversion equation was derived by associating the reflectance curve measured by the method of the present invention with the same sample position with the reflectance curve measured by VAR. The obtained conversion equation is as follows. θ = 51.122 + (P-260) × 0.035214 + (P-255) 3 × 6.9
× 10 ⁻⁸ Here, the third term is added to correct the influence of lens aberration. The range that can be correctly converted by the above equation is θ
= 43 to 60 ° (pixel numbers 47 to 488), and the data in this angle range is used for the film thickness analysis.

In this method, the reflectance data obtained from the sample is divided by the reflectance data obtained using only the substrate in advance. Thereby, it is possible to suppress the influence of the intensity distribution unevenness of the incident light and the noise caused by the optical system.

Next, the film thickness dependence of R (θ) will be described. Here, the sample 11 is made of a single-layer film of SiO ² (n = 1.46), and the substrate supporting the sample 11 is Si (complex refractive index).
n = 3.85-i0.024), and the incident angle is 35 ° to 65 °.
FIG. 3 shows the result of theoretical calculation of a reflectance curve obtained when the incident light wavelength was 632.8 nm and the thickness of the SiO ₂ film was changed between 0 and 1000 nm. However, it is divided by the reflectance of the substrate as in the case of the above measurement. a) is
s-polarized light and b) show the case of p-polarized light. It can be seen that when the film thickness is small, the period of change of the reflectance curve is long and the fluctuation width is small, and as the film thickness is large, the fluctuation period is short and the fluctuation width is large. Further, it can be seen that the period of change does not change between p-polarized light and s-polarized light, but the absolute value of the amount of change is larger for s-polarized light.

[0029]

EXAMPLES Next, specific examples of the present invention will be described. (Example 1) Wedge-shaped SiO ₂ film on Si A thick SiO ₂ film was formed on a Si wafer, immersed in an etching solution, and pulled up at a constant speed from the SiO ₂ film. An attempt was made to measure the film thickness distribution using a wedge-shaped SiO ₂ film 11 having a linear film thickness distribution. FIG. 4 shows an example of measurement of a reflectance distribution image near a film thickness of 600 nm. However, the data of 480 pixels in the y direction × 441 pixels in the θ direction is reduced to に 対 し in the y direction and is displayed in gray scale (0 (black) to 225 (white)). In the case of the s-polarized light in FIG. 4A, the position of the reflectance minimum point shifts due to the difference in the film thickness at each position in the y direction, which appears as a black line running obliquely. In contrast, FIG.
In the case of the p-polarized light shown in FIG. 3B, the fluctuation range of the reflectance curve is small, which coincides with the result shown by the theoretical calculation in FIG.

FIG. 5 shows a case where the film thickness distribution obtained from the reflectance distribution image is s-polarized light. Measurement line is slit length direction y
The three film thickness distributions obtained by moving the sample in parallel at intervals of 0.75 mm on the sample without changing the range of are also shown. For comparison, the results measured using BPR are shown by Δ. All results are in good agreement. On the other hand, no coincidence was obtained in the case of p-polarized light. This is presumably because the accuracy of the fitting was reduced due to the smaller variation range of the reflectance curve.

Next, FIG. 6 shows three film thickness distributions obtained by translating the measurement line in the same manner as before using s-polarized light near the film thickness of 1100 nm. By measuring line or BPR
And slightly different results were obtained. This can be explained by the fact that the measurement portion is the end of the sample, where the film thickness distribution produced is expected to be not linear but vary from place to place.

(Example 2) Spin coat Tefl on BK7
FIG. 7 shows the film thickness distribution of the sample 11 obtained from a Teflon film (n = 1.30) having a non-uniform film thickness formed on the BK prism (n = 1.52) by the spin coating method. Film thickness 1290
A film thickness distribution with a change in slope of 11340 nm was obtained.

(Example 3) Uniform positive resist film and uniform SiO ₂ film on Si A positive resist film (n
FIG. 8 shows a reflectance distribution image measured using s-polarized light for (= 1.62, d = 1000 nm). Since the film thickness is constant, the reflectance curves at the sample positions in the y direction are the same, and the reflectance does not change in the y direction. The obtained film thickness distribution is 974.
42 ± 0.12 nm, and the variation in the results was within ± 0.012%. In addition, as a result of repeating the measurement five times, 0.039%
There was reproducibility within.

(Conclusion) In each of the above embodiments, the film thickness distribution could be measured with a reproducibility of 0.039% or less with a data acquisition time of 1/30 second without mechanical scanning. The following means can be considered for further improving the accuracy and expanding the measurement range. (1) In order to increase the amount of information by increasing the incident angle width, the cylindrical lens 17 that forms the convergent light is changed to one with a high NA. (2) The accuracy of the calibration formula from the pixel number to the incident angle is increased using the calibration standard sample. (3) Raise the S / N of the reflectance measurement by using a lens with a reflection film prevention.

[0035]

As described above, according to the present invention, the thickness distribution of a thin film sample can be easily obtained along a straight line without performing mechanical scanning of light.

[Brief description of the drawings]

FIG. 1 is a schematic diagram showing an apparatus for measuring a film characteristic value distribution according to the present invention.

FIG. 2 shows various quantities for determining the reflectance.

FIG. 3 is a diagram showing reflectance in the case of s-polarized light and p-polarized light.

FIG. 4 is a diagram showing a reflectance distribution image in the case of s-polarized light and p-polarized light.

FIG. 5 is a view showing a SiO2 film thickness distribution (around 600 nm) obtained from an s-polarized reflectance distribution image.

FIG. 6 is a view showing a SiO2 film thickness distribution (around 1100 nm) obtained from an s-polarized reflectance distribution image.

FIG. 7 is a view showing a film thickness distribution of a Teflon film obtained from an s-polarized light reflectance distribution image.

FIG. 8 is a view showing a reflectance distribution image of a positive resist film measured using s-polarized light.

[Explanation of symbols]

 Reference Signs List 10 Measurement device for film characteristic value distribution 11 Thin film sample 12 Thin film sample holder 13 Light source 14 Polarizer 17 Cylindrical lens 20 CCD camera 21 Computer

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Katsuichi Kitagawa 1-1-1, Sonoyama, Otsu-shi, Shiga F-term in the Shiga Plant of Toray Industries (reference) 2F065 AA30 BB22 BB25 CC02 CC19 CC31 DD06 FF04 FF50 GG05 HH04 HH05 HH09 JJ03 JJ26 LL08 LL12 LL30 LL33 LL36 QQ00 QQ24 SS02 UU01 UU05 2G059 AA02 AA03 BB10 EE02 EE05 FF01 GG01 HH02 JJ11 JJ19 JJ20 KK04

Claims (6)

[Claims] 1. A step of irradiating a thin film sample with convergent slit light passing through a cylindrical lens, and making the reflected light from the sample coincide with one axis in the major axis direction of the convergent slit light and the other axis as incident angle characteristics. A step of detecting with a two-dimensional image sensor matched with the detection direction; and a reflectance R relating to a position y on the slit and an incident angle (θ) to the sample based on information from the two-dimensional image sensor.
(Y, θ), sample thickness (d), sample refractive index (n),
Obtaining a thickness of the sample along the long axis direction of the convergent slit light by fitting the theoretical reflectance RT according to the incident angle (θ) and at least the sample. Measuring method. 2. The film characteristic value distribution according to claim 1, wherein, when the thickness along the major axis direction of the convergent slit light is determined, the refractive index along the major axis direction of the convergent slit light of the sample is determined at the same time. Measuring method. 3. A thin film sample holder for holding a thin film sample, a light source, a cylindrical lens for converging light from the light source and projecting convergent light onto the thin film sample on the thin film sample holder, A two-dimensional image sensor that receives reflected light from the sensor and makes one axis coincide with the long axis direction of the convergent light, and the other axis coincides with the incident angle dependent characteristic detection direction, and information from the two-dimensional image sensor. Based on the position y on the slit
And the reflectance R (y, θ) according to the incident angle (θ) to the sample
, The thickness of the sample (d), the refractive index of the sample (n), and the incident angle (θ)
And at least the thickness d along the long axis direction of the convergent slit light of the sample by fitting the theoretical reflectance RT of
and a calculation unit for calculating (y). 4. The method according to claim 1, wherein the calculating section calculates the refractive index n (y) of the sample along the long axis direction of the convergent slit light when obtaining the thickness of the sample along the long axis direction. Item 3. A measuring device for film characteristic value distribution according to item 3. 5. The light source is a linearly polarized laser, and a polarization direction switching unit for switching the polarization direction of convergent slit light from the cylindrical lens between P-polarized light and S-polarized light is provided between the light source and the cylindrical lens. 4. The apparatus for measuring film characteristic value distribution according to claim 3, wherein: 6. The apparatus according to claim 5, wherein the polarization direction switching unit is formed of a polarizer or a wave plate.