JPH07229730A - Method for measuring thickness of amorphous coating film - Google Patents

Abstract

PURPOSE: To calculate the thickness of an amorphous coating film by selecting a take-off angle of photoelectron emission from a substrate and by calculating based on the take-off angle and a ratio of photoelectron emission from the substrate and photoelectron emission from the amorphous coating film. CONSTITUTION: The thickness of a thin oxide film 1 which is an amorphous coating film is calculated from Si in a substrate 2 and a ratio of Si2p inner shell photoelectron emission intensities Io and Ie against each of oxidized Si in the oxide film. These intensities are obtained by integrating XPS(X-ray photoelectron spectrometry) spectrum and by approximately subtracting background with linear interpolating between points of low binding energy side and high binding energy side with a spectral curve. Because of photoelectron forward scattering or photoelectron diffraction, dependency of photoelectron emission strength fluctuation according to take-off angle appears. When a take- off angle dependency of the photoelectron emission strength from the substrate 2 is measured in advance, and a measured thickness is calibrated, the exact thickness of an amorphous coating film on the crystal substrate 2 can be obtained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention generally relates to a method for measuring the thickness of a thin film, and more particularly to a method for measuring an amorphous film on a crystalline substrate.

[0002]

2. Description of the Prior Art The minimum dimensions of modern VLSI's have been reduced well below submicron. In the future, metal-oxide-semiconductor (MOS) field effect transistor (FET) gate oxide requirements will be less than 6 nm.
The native oxide thickness becomes comparable to the gate oxide thickness. Therefore, it is essential to accurately measure the thickness of the native oxide film before oxidation. Heretofore, the thickness of the oxide film has been measured by an ellipsometric method which is highly accurate and easy to operate. However, for oxide films less than about 2 nm thick, ellipsometry is not the most reliable measurement method.
For oxide films thinner than 3 nm, X-ray photoelectron spectroscopy (XP
S) is about to be accepted as a more reliable measurement method (F. J. Himpsel et al., “Phy.
s. Rev. B38 (1988) 6084 ").

In XPS analysis of a thin oxide film, it is most common to calculate the thickness by using the ratio of the photoelectron emission from Si atoms in the oxide film to the photoelectron emission from the Si substrate. Conventionally, an electron escape angle (escape angle) or a take-off angle (tak) of a photoelectron emission intensity from a crystalline substrate.
The dependence on e-off angle) was not taken into account in this type of calculation. On the other hand, studies on the peaks and valleys observed in the angle-dependent curves of photoemission intensity from single-crystal GaAs substrates and from atomic layers of various metals on clean surfaces and from adsorbed molecules have shown some recent studies on Si substrates. Reported with the report (L.
Kubler et al., "Surf. Sci. 251/252.
(1991) 305 "). The origin of the angle dependence is attributed to a phenomenon known by the name of either photoelectron diffraction or electron focusing or forward scattering.

[0004]

The present invention recognizes that the phenomenon of the angle dependence is an error factor that affects the thickness measurement by XPS, and 1) the thickness measurement by XPS is performed. The aim is to eliminate the risk of bad tolerances for misalignment of samples in the sample, and 2) to minimize the error resulting from this phenomenon.

That is, the present invention is a method for measuring the thickness of an amorphous film on a substrate, in which the photoelectron emission takeoff angle from the substrate is selected, and the above θ, the photoelectron emission from the substrate and the amorphous state are selected. Calculating the thickness d of the amorphous coating based on the ratio of photoemission from the coating.

[0006]

EXAMPLE A measuring method according to an example of the present invention will be described. The substrate actually used for this measurement was (001) -oriented with a native oxide film formed in the atmosphere after a series of chemical pretreatments including the well-known SCl cleaning and the removal of the oxide film by a diluted HF solution. It was cut from a Si wafer. It is degreased and washed in an ultrasonic bath with acetone, ethanol and ultrapure water, then dried with pure nitrogen,
It was loaded into the analysis room of the XPS instrument. Sample size is 2
It is 0 × 20 × 0.6 mm ³ .

The X-ray photoelectron spectrometer used for this work is M
Perkin Elmer PHI 5500 Multitechnic with gKα / AlKα source and monochromated AlKα source
System. In this measurement, the MgKα source was operated at 250 W with the electron acceleration voltage set to 12 kV. Energy resolution is 1.22 eV for Si2p _3/2 peak
Of the full width at half maximum (FWHM) of The geometrical positional relationship between the X-ray source and the opening of the electron energy analyzer is fixed, with respect to the plane defined by the analysis axis of the X-ray source and the electron energy analyzer (hereinafter referred to as the analysis plane). The sample stage was rotated around a vertical axis to allow polar angle scanning of photoemission. The cleaved side of the substrate is equivalent to the crystal

[Outer 1] Parallel to the plane, but aligned or at 45 ° to the same plane. Angular scanning was performed with a sample manipulator by changing the takeoff angle defined by the direction of photoemission and the plane of the surface of the substrate. The takeoff angle was calibrated at 90 ° due to the strong photoemission peaks from the <001> substrate and the takeoff angle was scanned from 10 ° to 100 °.

The optics of the electron energy analyzer were set so that the nominal conical acceptance half-angle was 2 °, which corresponds to an angular resolution of about 4 °. The measurement by XPS is
It was carried out at a background pressure lower than 1 × 10 ^-9 Torr.

As shown in FIG. 1, the thickness of the thin oxide film (1), which is an amorphous film, is determined by the Si2p core photoelectron emission intensity with respect to Si in the substrate (2) and oxidized Si in the oxide film. Calculated from the ratio of Io and Ie. These intensities are X
It is determined by integrating the PS spectrum and approximately subtracting the background by linear interpolation between the low and high binding energy sides of the spectral curve and the points that are the troughs of the curve. The formula used to calculate the thickness d is shown below (takeoff angle dependence will be described later).

[Equation 1] Here, Ke and Ko are spectrometer functions, σe and σo are atomic photoionization cross sections, No and Ne are Si atom density per unit volume, λo and λe are electron escape depths for Si and SiO ₂ , respectively, and θ is Takeoff angle. The value used for σo / σe is 1.1, and λo and λe are 2. respectively for the Si2p core photoelectrons excited by the MgKα line.
5 nm and 2.3 nm. The ratio Ke / Ko is assumed to be 1.

The apparent thickness of native oxide on Si wafers, measured by XPS, has been shown to be a strong function of crystallographic orientation. The angle-dependent curve of photoelectron intensity from a crystalline Si wafer has large peaks and valleys, whereas the angle-dependent curve for photoelectron emission from an amorphous oxide film was not. This difference explains the angular dependence of the oxide film thickness as measured by XPS. The strong angular dependence of photoemission from crystals is attributed to a phenomenon known as photoelectron forward scattering or photoelectron diffraction.

FIG. 2 shows XP as a function of takeoff angle.
The thickness of the oxide film measured by S is shown. Angular scan is equivalent to board

[Outside 2] Performed on the analysis plane adjacent to one of the planes. The scanning step was 2.5 °.

Instead of the expected constant thickness value, a thickness variation of a few nanometers was observed in this range of takeoff angles. This is because when a particular takeoff angle point is chosen where the curve has a strong dependence, misalignment of the takeoff angles of the samples can introduce this degree of measurement error into the comparative measurements between samples. It means that there is.

FIG. 3 shows a comparison of the Ie and Io angle-dependent curves for the same analysis plane as previously described. The Ie curve shows peaks and troughs, which locally fluctuate up to about 30% and are symmetrical with respect to 90 °, that is, the <001> axis of the substrate, but Io has a gentle angular dependence. Is shown. The control for this feature is the origin of the thickness variation seen in FIG.

This contrast can be attributed to the phenomenon known as photoelectron diffraction or electron focusing or forward scattering. At this time, photoelectron emission from the atom is strengthened in the direction of the closest atom. The enhancement of the photoelectron emission depending on the angle is observed for the photoelectron emission Ie from the Si substrate which is a single crystal, but is not observed for the photoelectron emission Io from the oxide film which is an amorphous. Therefore, the peaks have crystal axes <111>, <112>, <113>, <1 as shown in FIG.
15> and <001>.
The peaks between 62.5 ° and 77.5 ° are 64.
This can be interpreted as the superposition of peaks for the <113> and <115> axes that are theoretically expected at 8 ° and 74.2 °.

FIG. 5 shows the takeoff angle dependence of Io and Ie scanned on the analysis plane adjacent to the (100) plane of the substrate. In the Ie curve,

[Outside 3] Several major peaks are seen at different angles to the Ie peaks scanned in the plane. this is,
It is shown that the nature of this phenomenon depends essentially on the crystallographic orientation of the substrate for the XPS device. The angular positions of the peaks are, as shown in FIG. 6, crystal axes <011>, <012
>, <013>, and <001> are in good agreement with those expected for the same phenomenon as described above. Broad peaks near 70 ° are theoretically expected to appear at 63.4 ° and 71.6 °, respectively <
It can be interpreted as a superposition of peaks on the 012> and <013> axes.

Another noteworthy feature of this Ie curve is that
It is a region with weak angle dependence in the range of 0 ° and 62.5 °. Since Io, unlike Ie, does not have a strong angle dependence, the ratio Io / Ie to this angular region is expected to have a weak angle dependence, and therefore the takeoff angle of the sample is matched in the thickness comparison measurement. It is expected to be more tolerant of outliers than other areas.

This argument is confirmed by the angle-dependent curve for thickness calculated on the basis of these curves, as shown in FIG. This curve shows a weak angle dependence in the region between 50 ° and 62.5 °, with a thickness variation of the order of 0.06 nm. This is a region unique to the {100} analysis plane,

[Outside 4] No equivalent angular width is observed in the curve to the plane. This can be a useful "analysis window" for this type of XPS thickness measurement.

As described above, usually, due to the photoelectron forward scattering or photoelectron diffraction, the dependence of the fluctuation of the photoelectron emission intensity depending on the takeoff angle appears. Therefore, it may be necessary to consider the influence of the crystal orientation here. However, if the take-off angle dependence of the photoelectron emission intensity from the substrate is measured in advance and the thickness measured using equation (1) is calibrated, the accurate thickness of the amorphous thin film on the crystalline substrate can be calculated. Is always obtained.

In the example described above, the measurement of thin silicon oxide on single crystal silicon was described. However, this idea can be applied to any combination of amorphous thin films on any crystalline substrate. For example, the amorphous thin film may be a high dielectric constant material such as silicon nitride, silicon oxynitride and thin ferroelectrics or tantalum pentoxide.

Further, the following matters will be disclosed.

(1) In the method of measuring the thickness of an amorphous coating on a substrate, the photoelectron emission takeoff angle from the substrate is selected, and the above θ, the photoelectron emission from the substrate and the amorphous coating are selected. The thickness d of the amorphous film based on the ratio to the photoelectron emission intensity of
A method of measuring the thickness of an amorphous coating comprising the step of calculating.

(2) The step of calculating the thickness d of the amorphous coating is

[Equation 2] The method according to (1), which includes the calculation of

(3) The substrate is Si (111) or Si
(100) and the amorphous coating is SiO ₂ (1)
The method described in.

(4) The method according to (1), wherein the step of selecting the photoelectron emission takeoff angle θ includes the step of measuring the dependence of the photoelectron emission intensity on the takeoff angle.

(5) The method according to (4), wherein the step of selecting the light emission take-off angle θ selects a take-off angle where the dependence is weak.

(6) The method according to (5), wherein the takeoff angle between 50 ° and 62.5 ° on the analysis plane {100} of the Si (100) substrate is selected.

[Brief description of drawings]

FIG. 1 is a diagram showing the measurement of the thickness of a thin oxide film formed on a Si substrate by XPS, which is an embodiment of the present invention.

[Fig. 2]

[Outside 5] The graph which shows the dependence of the thickness of an oxide film with respect to the takeoff angle scanned by the plane.

[Figure 3]

[Outside 6] The graph which shows the dependence of Ie and Io with respect to the takeoff angle scanned by the plane.

[Figure 4]

[Outside 7] Schematic diagram showing the principal axes of enhancement of photoemission in the plane.

FIG. 5: Ie for takeoff angle scanned in (100) plane
And a graph showing the dependency of Io.

FIG. 6 is a schematic diagram showing the principal axes of enhancement of photoemission in the (100) plane.

[Figure 7]

[Outside 8] The graph which shows the dependence of the thickness of an oxide film with respect to the takeoff angle scanned by the plane.

[Explanation of symbols]

 1 Amorphous film 2 Substrate θ Photoemission from the substrate Takeoff angle

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Tadahiro Yoneda 17 Miyukigaoka, Tsukuba, Ibaraki Prefecture, Texas Instruments Tsukuba R & D Center (72) Inventor Yashiro Nishioka 17 Miyukigaoka, Tsukuba, Ibaraki Texba Instruments Tsukuba R & D Center

Claims (1)

[Claims] 1. A method for measuring the thickness of an amorphous coating on a substrate, wherein a photoelectron emission take-off angle from the substrate is selected, and the above θ and the photoelectron emission from the substrate and the amorphous coating are selected. A method of measuring the thickness of an amorphous coating comprising the step of calculating the thickness d of the amorphous coating based on the ratio with photoemission.