JP3258852B2 - Abnormal detection method for dry etching equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for measuring an etching selectivity in dry etching used for manufacturing a semiconductor device.

[0002]

2. Description of the Related Art Conventionally, in order to know the etching selectivity in dry etching using a plasma apparatus, it is necessary to use:
A method of measuring and obtaining an actual rate and a method of SEM observation of a cross section of a sample wafer have been used.

[0003]

However, it takes time and effort to measure the etching rate under a large number of conditions, so that the number of conditions under which the measurement can be performed is limited. Sometimes made up for the number. However, in this case, it is difficult to predict the process parameters unique to the apparatus, and it is difficult to judge them comprehensively as the process parameters increase. Further, when the point happens to be a peculiar point, the parameter may need to be optimized again. The method of performing cross-sectional SEM observation also has a problem that the method is easily affected by the previous process.

[0004] In view of these problems, an object of the present invention is to provide a method capable of setting an etching selectivity by a simple method and confirming the presence or absence of an abnormality in the apparatus.

[0005]

According to the present invention, there is provided a dry etching method using emission spectroscopy.
In dry etching of a silicon-based material using a gas containing oxygen and halogen as a process gas, an intensity ratio between emission of a specific wavelength of oxygen radicals and emission of a specific wavelength due to halogen radicals is obtained by emission spectroscopy, and the intensity ratio is used. The etching selection ratio is set. Further, bromine or chlorine is used as the halogen. Further, in the dry etching method, the wavelength to be measured is 777 nm (oxygen) and 78 nm.
0 nm (bromine) or 777 nm and 808 nm (chlorine).

[0006]

The relative emission intensity of the oxygen plasma greatly depends on the etching selectivity. This is because the dissociation energy of oxygen gas is relatively higher than the dissociation energy of halogen, so that the relative intensity of oxygen increases in a region where the plasma density or the electron density is high, and the selectivity also increases at that time. . Therefore, by measuring the relative intensity ratio of light emission due to oxygen radicals, the etching selectivity can be easily estimated. Further, even when an abnormality occurs in the device, a change in the electron density occurs. Therefore, the abnormality in the device can be known by measuring the emission intensity. Furthermore, since the distribution of the electron density in the plasma also appears as a change in the emission intensity, the uniformity of the plasma can be evaluated from the emission intensity.

[0007]

The present invention will be described below with reference to examples.

FIG. 1 shows an ECR plasma etching apparatus for a 6-inch wafer used in the present invention. This device generates a microwave of 2.45 GHz from a magnetron and introduces it into a vacuum chamber using a waveguide. A magnetic field condition is set in the chamber so as to satisfy electron cyclotron resonance conditions by an external coil, and the chamber is evacuated using a turbo molecular pump.

With this apparatus, the etching conditions for the polysilicon film used for the resist mask will be examined. Using a mixed gas of hydrogen bromide, chlorine, and oxygen as the process gas, changing the process parameters of the device (pressure, magnetic field, RF power, microwave power, gas flow rate, gas mixing ratio, etc.) The emission intensity ratio (specifically, the emission intensity ratio at 777 nm and 780 nm) due to the oxygen / bromine radical is determined.

The emission intensity is measured as follows.
The light emission of the target plasma is collected through a measurement window provided on the side wall of the processing chamber using a lens system, and is introduced into a spectroscope using a diffraction grating through an optical fiber. Then, the intensity of the light dispersed by the spectroscope is measured by an area sensor using a CCD. The emission intensity ratio is calculated from the peak ratio of the emission at each wavelength. FIG. 2 shows the obtained results.

In FIG. 2, the vertical axis represents the selection ratio, and the horizontal axis represents the emission intensity ratio of oxygen / bromine radicals. The selection ratio on the vertical axis is obtained as follows. The etching rate of the polysilicon and oxide film under each condition is measured using a wafer for rate measurement. The wafer that measures the rate
For polysilicon, 100nm film thickness on silicon wafer
A silicon oxide film is formed by thermal oxidation, and a film obtained by depositing polysilicon by 500 nm by CVD thereon is used. For an oxide film, a wafer having a thermal oxide film formed on a wafer by 500 nm is used. The polysilicon is etched for 1 minute and the oxide film is etched for 2 minutes to determine the change and rate of the film thickness before and after that.
The selectivity is calculated by dividing the polysilicon rate by the oxide film rate.

The upward triangle in FIG. 2 indicates the height of the ECR resonance region (magnetic field intensity is 875 G) as viewed from the wafer by 110.
mm to 200 mm, diamonds indicate pressures changed between 0.77 Pa to 3 Pa, and downward triangles indicate magnetic field gradients in the ECR resonance region of 30 G / mm to 70 mm.
G / mm, the black circles indicate the amount of oxygen gas
The squares are obtained by changing the power of the microwave from 700 W to 1200 W.

FIG. 2 shows that a correlation is established between the emission intensity ratio and the selection ratio. This shows that the etching selectivity can be set more efficiently than the emission intensity ratio by first measuring some points of the selectivity. In the actual setting of the etching conditions, it is often sufficient that the selectivity is higher than the target value. Therefore, this degree of correlation is sufficient.

Although polysilicon is used as the resist mask here, the measurement can be performed in the same manner when amorphous silicon or crystalline silicon is used. When doping with impurities, the value of the selectivity itself is different, but the same tendency can be obtained. According to the present invention, since changes due to a plurality of parameters can be predicted, optimization can be easily performed even when parameters that can be changed are limited due to a limitation of hardware of the apparatus or when parameters are not independent. The device is a normal RI
In the case of the E apparatus, the range of process parameters that can be changed is limited, but optimization can be performed similarly.

The wavelengths limited here are 777 nm and 78
Although the wavelength of 0 nm is selected, any wavelength may be selected as long as it is a wavelength of light emission by oxygen radicals and halogen radicals. These two wavelengths have the advantage that noise due to other light emission is small, and because the wavelengths are close to each other, they can be measured simultaneously when emission spectroscopy is performed by an emission spectrometer using a CCD element. However, when configuring a measuring device using a filter, it is better that the wavelengths are separated,
For example, wavelengths of 777 nm and 808 nm are used.

Next, FIG. 3 shows the relationship between the ratio of the light emission intensity and the pattern shift for the same measurement result. When the O / Br ratio is equal to or more than a certain value, it can be seen that the pattern shift hardly occurs. From this, it can be seen that it is only necessary to make the emission intensity ratio larger than a certain value in order to solve the problem (dependence of pattern width on density) of the deposit.

Subsequently, the same experiment is performed using a commercially available helicon etcher. The rate measurement was performed in the same manner as in the ECR etching apparatus. For emission spectroscopy, a filter type is used, and the ratio of the emission intensity using a filter transmitting a wavelength of 777 nm to the emission intensity using a filter transmitting a wavelength of 808 nm is measured. In this device, the pressure and the power of the helicon wave are parameters that can be changed.
The result of this experiment is shown in FIG. The horizontal axis is the emission intensity ratio (oxygen / chlorine), and the vertical axis is the selection ratio. As shown in FIG. 4, it can be seen that a positive correlation is established between the emission intensity ratio and the selection ratio. From the above, it can be seen that the selectivity can be easily obtained by the method of measuring the emission intensity ratio before etching.

Next, a case where an integrated circuit is manufactured using an etching apparatus will be described. Usually, when etching a wafer, discharge is performed using one or more dummy wafers to stabilize the discharge state in the chamber. By measuring the light emission during discharge of the first dummy wafer and obtaining the light emission intensity ratio, it can be determined that the selection ratio has not deteriorated. In addition, the discharge state changes due to deterioration of the pumping characteristics of the pump, deterioration of the microwave generator, failure of the mass flow controller, and the like. Since these influence the light emission intensity ratio, abnormality of the device can be detected by measuring the light emission intensity ratio and comparing it with a normal value. During the discharge of the dummy wafer, the EPD (end point detection device) is unnecessary, so that the measurement can be easily performed by modifying the EPD device so that the intensity ratio of the light emission can be obtained.

Even in an apparatus in which oxygen gas is not used during actual dry etching, the emission intensity when normal operation is performed by introducing plasma by introducing oxygen gas and halogen gas, or oxygen gas and argon gas. Abnormality determination can be easily performed by comparing with the ratio. Examples of gas combinations in this case include any combination of oxygen and nitrogen, such as oxygen and nitrogen, carbon dioxide and chlorine, oxygen and hydrogen bromide, or oxygen and argon, oxygen and nitrogen, A combination of oxygen and helium may be used. However, since oxygen and helium have close emission peak intensities, spectroscopy is difficult, and the combination of oxygen and argon facilitates measurement.

Further, by measuring the plasma at a plurality of locations, the uniformity of the plasma at the time of etching can be evaluated. The luminescence of various parts of the plasma is collected using an optical fiber as shown in FIG.
The uniformity of the plasma is evaluated by determining the emission intensity ratio of each part. FIG. 6 is a view showing a state 100 above the wafer using the apparatus of FIG.
The emission intensity ratio was measured at a height of mm in the radial direction of the wafer. The horizontal axis represents the position, and the vertical axis represents the emission intensity ratio. If you simply measure the absolute intensity of light emission,
Although it is affected by the plasma shape, the measurement distance, and the like, if the relative intensity ratio is measured, it is difficult to be affected by the plasma shape, so that the uniformity of the plasma can be accurately evaluated.

[0021]

According to the present invention, the etching selectivity in dry etching can be measured by a simple method, and abnormality of the apparatus can be easily detected even during manufacturing.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of an ECR plasma etching apparatus used in the present invention.

FIG. 2 is a diagram illustrating a relationship between a light emission intensity ratio of oxygen and bromine radical and an etching selectivity.

FIG. 3 is a diagram showing a relationship between a light emission intensity ratio of oxygen and bromine radical and a pattern shift.

FIG. 4 is a diagram showing the relationship between the emission intensity ratio of oxygen and chlorine radicals and the etching selectivity in a helicon wave etcher.

FIG. 5 is a diagram illustrating a configuration of a uniformity measuring device used in the present invention.

FIG. 6 is a diagram showing the result of measuring the emission intensity ratio of oxygen and bromine in the radial direction of the wafer.

──────────────────────────────────────────────────続 き Continuation of front page (56) References JP-A-63-244847 (JP, A) JP-A-4-242927 (JP, A) JP-A-62-45116 (JP, A) JP-A-1- 179326 (JP, A) JP-A-5-291188 (JP, A) JP-A-5-67665 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) C23F 4/00 H01L 21 / 3065

Claims (3)

(57) [Claims] 1. Process of processing a gas containing oxygen and halogen
Dry etching of silicon material used as gas
In the dry etching equipment
Oxygen radical emission at a specific wavelength and halogen radical
The intensity ratio of light emission of a specific wavelength
And the value of the emission intensity ratio of the dry etching system in normal operation.
By comparing, the abnormality of the dry etching apparatus
Of dry etching equipment characterized by detecting
Always detection method. 2. A bromine or halogen as the halogen according to claim 1.
Is a driver using emission spectroscopy, which uses chlorine.
An abnormality detection method for etching equipment. 3. A method for determining an emission intensity ratio according to claim 1.
, The wavelength to be measured is 777 nm (oxygen) and 780
nm (bromine) or 777 nm and 808 nm (salt
Dry etching using emission spectroscopy
An abnormality detection method for a switching device.