JPH08195385A - Dry etching method - Google Patents

Abstract

PURPOSE: To realize a vertically etched form and to lessen an isolated line pattern and an inner line pattern in dimensional difference between them by a method wherein the gas pressure, the gas ratio, the gas discharge, the temperature of a specimen stage, and a bias power are set optimal in combination in a chamber. CONSTITUTION: When an inner line pattern is smaller than an isolated line pattern, and a line pattern is larger than a resist pattern in line width, at least one of the gas pressure, the gas discharge amount of the vacuum chamber, the frequency and power of the high-frequency electric power, the side wall protecting gas ratio, and the temperature of the specimen stage is changed. By this setup, the line pattern is enhanced in etching rate, and the inner line pattern is set lower than the isolated line pattern in etching rate. Therefore, the inner line pattern approximates to the isolated line pattern in line width, and the line pattern approximates to the resist pattern in line width.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a dry etching method using plasma.

[0002]

2. Description of the Related Art As a dry etching method, a sample stage which is a cathode electrode is installed inside a chamber of a plasma generator, and high frequency power is applied to the sample stage to form a self DC bias. It is known that a sample to be etched formed of a substrate or a film formed on the substrate is dry-etched mainly or auxiliaryly by accelerating and inducing ions toward a sample stage. As described above, the plasma using the high frequency discharge is applied to the dry etching process for fine processing.

Advances in modern high-density semiconductor integrated circuits are bringing about changes comparable to the industrial revolution. Higher densities have been realized by miniaturization of device dimensions, improvement of devices, and enlargement of chip size. The miniaturization of device dimensions has advanced to about the wavelength of light, and the use of excimer lasers and soft X-rays is being considered for lithography.
Along with lithography, dry etching and thin film formation play an important role in realizing fine patterns.

Dry etching applied to fine processing will be described below. The dry etching is a technique of removing an unnecessary portion of the sample to be etched by utilizing a chemical or physical reaction between radicals and ions generated by plasma and a solid phase surface of the sample to be etched. Reactive ion etching (RIE), which is the most widely used dry etching technique, is that when an etching target sample is exposed to a high-frequency discharge plasma of an appropriate gas, unnecessary parts of the etching target sample are removed by the etching reaction. Is. The required portion is usually protected by the photoresist pattern used as a mask.

In the following description, the pattern formed on the sample to be etched when the sample to be etched is dry-etched using the resist pattern as a mask is particularly called a line pattern.

In the dry etching technique, formation of a substantially vertical etching shape according to a mask pattern of fine dimensions is performed by forming a line pattern (which is located inside of a line pattern group consisting of a plurality of line patterns formed close to each other). Hereinafter, it is also realized for an internal line pattern), an outermost line pattern in the line pattern group (hereinafter referred to as an external line pattern), and an isolated line pattern formed separately from the line pattern group. In addition, it is required to reduce the dimensional difference between the isolated line pattern (or the external line pattern) and the internal line pattern as much as possible.

As a countermeasure against these requirements, conventionally, by lowering the gas pressure inside the chamber and increasing the degree of vacuum, ions are transported to the sample stage while being accelerated in the sheath region formed near the sample. In addition, while minimizing the scattering due to the collision between the ions and neutral particles, in order to prevent the etching to the side wall of the line pattern due to obliquely incident ions that always exist in a certain proportion,
A method of adding a sidewall protecting gas that generates radicals responsible for sidewall protection has been adopted.

[0008]

As described above, the etching method adopting a low degree of vacuum and the addition of a gas for protecting the sidewall can obtain higher perpendicularity of the incident angle distribution of ions to the sample surface, To some extent, it is possible to control the etching protection of the side wall of the line pattern by obliquely incident ions, and the formation of a substantially vertical etching shape according to the mask pattern size is individually performed for the isolated line pattern (or the external line pattern) or the internal line pattern. Can be realized. However, the isolated line pattern,
Simultaneously realize the formation of an etching shape perpendicular to both the external line pattern and the internal line pattern, and sufficiently reduce the dimensional difference between the isolated line pattern (or external line pattern) and the internal line pattern. It was difficult to do.

In view of the above, according to the present invention, the gas pressure inside the chamber, the gas ratio, the gas exhaust amount, the temperature of the sample stage and the bias.
By optimizing the combination of powers, the energy and angular distribution of ions near the surface of the sample stage and the ratio of radicals that play a role in sidewall protection can be controlled, and these can be used for isolated line patterns, external line patterns, and internal line patterns. The object of the present invention is to realize the formation of a vertical etching shape and to sufficiently reduce the dimensional difference between the isolated line pattern (or the external line pattern) and the internal line pattern.

[0010]

The present invention is directed to the gas pressure of a raw material gas introduced into a vacuum chamber, the discharge amount of gas discharged from the vacuum chamber, and the high frequency power applied for forming a self-bias in the vacuum chamber. When at least one parameter of the parameter group consisting of the frequency, the power of the high frequency power, the ratio of the sidewall protecting gas in the raw material gas, and the temperature of the sample stage is changed, the sidewall protecting radicals are deposited on the sidewall of the line pattern. And the amount of etching of the deposited side wall protective film due to the ions,
This finding was made based on the finding that the isolated line pattern, the external line pattern, and the internal line pattern change independently.

Hereinafter, in the dry etching method,
Values of external operating parameters such as gas pressure, gas ratio, gas exhaust rate, bias power, frequency of high frequency power, and sample stage temperature, and isolated line pattern of the sample to be etched on the sample stage, internal line pattern sidewall and external line Percentage of side wall protective radicals flying to each side wall of the pattern,
Side wall protection Relation between radical adsorption rate, ion flux, ion angular distribution, ion energy distribution and plasma internal parameter values, and side wall protection deposited on each side wall of isolated line pattern (or external line pattern) and internal line pattern The relationship between the film deposition amount and the etching amount will be described.

In the following description, the contents of comparing the isolated line pattern and the internal line pattern also apply to the comparison of the outer surface of the external line pattern and the internal line pattern.

(A) First, the above-mentioned relationship centered on the value of an external operating parameter called gas pressure will be described.

(1) Side wall protection deposited on the side wall of the line pattern by the side wall protection radicals that play a role of side wall protection such as reaction product radicals existing in the chamber and radicals generated from the side wall protection additive gas. The amount of film is larger in the isolated line pattern than in the internal line pattern regardless of the gas pressure in the chamber. The reason for this is that the expected solid angle of the side wall protective radical flux that is incident isotropically from above the sample to be etched is approximately π / 2 in the case of the isolated line pattern side wall.
It is smaller than π / 2 in the case of the inner line pattern side wall, and is smaller than π / 2 especially when the pattern aspect ratio (the height of the line of the line-and-space pattern divided by the space width) is large. This is because it will be considerably smaller.

When the ratio of the side wall protective radicals is reduced, the amount of the side wall protective film deposited on the side wall of the isolated line pattern and the amount of the side wall protective film deposited on the side wall of the internal line pattern decrease while maintaining a constant ratio. . However, the reduction amount of the sidewall protection film with respect to the reduction amount of the sidewall protection radical is larger in the isolated line pattern sidewall than in the internal line pattern sidewall. That is, the amount of change in the film thickness of the sidewall protective film on the sidewall of the isolated line pattern is larger than the amount of change in the film thickness of the sidewall protective film on the sidewall of the internal line pattern.

(2) Next, the behavior of incident ions will be described.

The incident ions are incident on the surface of the sample while being accelerated in a direction substantially perpendicular to the surface of the sample table by the electric field in the direction perpendicular to the surface of the sample table provided in the sheath region. However, due to the collision with the neutral particles in the sheath region, the angular component is scattered to some extent and enters the surface of the sample.

Ions obliquely incident on the surface of the sample, after removing the side wall protective film on the side wall of the line pattern from the side wall of the line pattern standing perpendicular to the surface of the sample, after etching. This has the effect of changing the pattern profile from forward taper to vertical or from vertical to inverse taper. The more perpendicular the incident angle of the ions is to the side wall of the line pattern, that is, the larger the scattering angle of the ions, the greater the etching ability with respect to the side wall protective film. On the contrary, ions incident perpendicularly to the sample surface have an incident angle substantially parallel to the side wall of the line pattern, so that the etching ability for the side wall protective film is small.

The above phenomenon is particularly remarkable on the side wall of the isolated line pattern because the oblique incident ion has a large expected solid angle and the incident ion directly collides with the side wall. On the other hand, on the side wall of the internal line pattern, the expected solid angle of the obliquely incident ions is small, and is particularly small when the aspect ratio of the line pattern is large.
Therefore, incident ions having a scattering angle of a certain size or more are reflected without entering the space portion of the line and space pattern. In other words, the ion angle distribution component of the incident ion having a large scattering angle cannot directly fly to the lower part of the side wall of the internal line pattern, and is more perpendicular to the sample surface and has a relatively smaller scattering angle. Only the ion angle distribution component can selectively fly. That is, of the ion flux having a wide ion angle distribution component, only the component having a certain scattering angle or less is collimated, and there is an effect that it is allowed to fly to the lower portion of the sidewall of the internal line pattern. Hereinafter, this effect will be referred to as "ion collimation effect".

Incident ions having a large scattering angle, which are somehow incident near the upper portion of the inner line pattern side wall, must be reflected several times between the inner line pattern side wall of the line pattern group and the opposite line pattern side wall, and the inner line pattern side wall is not reflected. Cannot penetrate into the bottom of the pattern side wall. During this time, the energy of incident ions is reduced and the ability to etch the deposited protective film is reduced. Also, the incident ion flux is reduced as compared with the isolated line pattern.

(3) The above phenomena will be summarized as follows. That is, the amount of change in the amount of deposited side wall protective film with respect to the change in the ratio of the side wall protective radicals in the chamber is greater in the isolated line pattern side wall than in the internal line pattern side wall. Due to the ion collimation effect, in the isolated line pattern, the effect of removing the side wall protective film by obliquely incident ions is greater than that in the side wall of the internal line pattern. The change in the ion incident angle appears as a large change in the etching amount of the side wall protective film on the side wall of the isolated line pattern.

As described above, the isolated line pattern is relatively sensitive to changes in the proportion of side wall protecting radicals and changes in the ion incident angle, and the internal line pattern is relatively insensitive.

(4) As described above, ions collide with neutral particles in the sheath region and enter the sample surface with angular components scattered to some extent, but the gas pressure is sufficiently low and high. In the vacuum region, the incident ions are less likely to collide with neutral particles in the sheath region, have many components that are incident relatively vertically, and are incident on the sample surface with a relatively small scattering angle overall. Come on. That is, in the high vacuum region, many incident ions are incident in a substantially uniform manner along the side wall of the line pattern. This effect will be called the "pseudo parallel beam effect".

In the case of such an incident ion angle distribution, the ion flux incident on the side wall of the isolated line pattern is larger than the ion flux incident on the side wall of the internal line pattern, but the difference is larger than in the medium vacuum region. There is no. That is, the difference in etching ability of the deposition protection film with respect to the isolated line pattern and the internal line pattern is not so different as in the medium vacuum region. Further, the angle of incident ions with respect to the side wall is relatively small, and the proportion of ions incident on the side wall of the line pattern at a large angle is small. Therefore, the ability to etch the deposition protection film is smaller than that in the medium vacuum region, and is particularly small on the side wall of the isolated line pattern.

(5) In the above description, the relation between the deposition amount of the side wall protection film and the etching amount of the obliquely incident ions in the deposition protection film has been described. When the etching amount of the protective film is larger than the deposition amount of the protective film, the profile of the line pattern of the sample to be etched becomes an inverse taper, and the dimension of the line pattern becomes smaller than the resist mask dimension. In this case, when the gas pressure is reduced to reduce the proportion of obliquely incident ions, the taper is changed from the inverse taper to vertical, and the reduction amount of the line pattern dimension of the sample to be etched is also alleviated.

(B) Next, the above-mentioned relationship centering on the combination of two external operating parameters of the gas pressure P and the exhaust amount Q will be described.

When the gas pressure P is increased, the following internal plasma parameter changes occur. As described above, the standard deviation σ of the ion scattering angle representing the spread of the incident ion angle distribution increases, and the ion energy E _i of the ions incident on the sample surface decreases. Since the increase of the gas pressure P means the increase of the raw material gas, the reactive radical flux F _R and the ion flux F _i are increased as long as the input power has the ability to sufficiently ionize and excite the raw material gas. As a result, the proportion of side wall protective radicals such as reaction products and sputtered resist in the chamber is increased, and the side wall protective radical flux F _RP is increased.

The ion energy E _i and the standard deviation σ of the ion scattering angle under the assumption of a normal distribution are internal parameters related to the etching ability of the sidewall protective film deposition amount, and the decrease of the ion energy E _i indicates the etching ability. Since the decrease and the increase in the standard deviation σ of the ion scattering angle result in the improvement in the etching ability, it cannot be simply determined whether the increase in the gas pressure P results in the improvement or the decrease in the etching ability. On the other hand, an increase in the side wall protective radical flux F _RP obviously causes an increase in the side wall protective film deposition amount. Therefore, it cannot be simply determined whether the increase of the gas pressure P causes the increase of the sidewall protective film deposition amount or the decrease thereof.

However, if the exhaust amount is increased and the ratio of the side wall protective radicals in the chamber is reduced with the gas pressure P kept constant, etching of the side wall protective film causes an increase in the film deposition amount. Be superior. That is,
The side wall protective film thickness can be controlled by controlling the exhaust amount.

(C) Next, the above-mentioned relationship centering on the combination of the two external operating parameters of the bias power W _B and the displacement Q will be described.

When the bias power W _B is increased, the following plasma internal parameter changes can be considered.
First, the energy E _i of the ions incident on the sample surface increases. The ion scattering angle standard deviation σ decreases. Further, an increase in the bias power W _B means an increase in ionization excitation of the raw material gas near the sample surface, and therefore an increase in the ion flux F _i and the reactive radical flux F _R. As a result,
The ratio of side wall protective radicals such as reaction products and sputtered resist in the chamber increases, and the side wall protective radical flux F _RP increases.

The ion energy E _i , the ion flux F _i and the ion scattering angle standard deviation σ are internal parameters related to the etching ability of the sidewall protection deposited film. The increase of the ion energy E _i and the ion flux F _i brings about the improvement of the etching ability. On the other hand, a decrease in the ion scattering angle standard deviation σ and an increase in the sidewall protection radical flux F _{RP result in} an increase in the sidewall protection film deposition amount. Therefore, it cannot be simply determined whether the increase of the bias power W _B causes the increase of the sidewall protection film deposition amount or the decrease thereof.

However, if the exhaust amount is increased and the ratio of the side wall protective radicals is reduced with the bias power W _B kept constant, etching of the side wall protective film increases the deposition amount of the side wall protective film. Will be superior to. That is, the thickness of the sidewall protective film can be controlled by controlling the exhaust amount.

(D) Next, the above-mentioned relationship centering on the combination of two external operating parameters of the frequency f of the high frequency power and the gas pressure P will be described.

That is, the sheath width d on the cathode side formed in the vicinity of the sample table is as follows.

D = K ₁ / (P ^m · f ⁿ ) ... (1) where P is gas pressure, f is frequency, and m is a positive real number, which is larger than about 1/3 and about 1/2. Smaller than. n is a positive real number, which is larger than about 1/2 and smaller than about 1. This is K. Harafuji, A. Yama
no and M. Kubota: Jpn. J. Appl. Phys. vol.33 (1994)
p2212, and N. Mutsukura, K. Kobayashi and Y. Machi:
Vol.68 (1990) p.2657, J. Appl. Phys.

The mean free path λ of the ions, which is mainly derived from the elastic collision scattering and the charge exchange scattering between the ions and the neutral particles, is inversely proportional to the gas pressure P and can be expressed as follows. .

Λ = K ₂ / P (2) Ions that have left the boundary between the bulk plasma region and the sheath region are scattered by collision with neutral particles in the sheath region while being transported to the cathode with the sample stage. The quantity η proportional to the probability of being

Η = d / λ (3) By substituting the equations (1) and (2) into the equation (3), η = d / λ = (K ₁ / (P ^m · f ⁿ )) _{× (P / K 2) =} (K 1 / K 2) × (P 1-m / f n) ~ (K 1 / K 2) × (P / f) 1/2 relationship is obtained. Here, K ₁ and K ₂ are constants, and means that they are almost equal. The following conclusions can be obtained from the above equations.

When the gas pressure P and the frequency f are increased,
From equation (1), the sheath width d, which is the distance traveled while the ions starting from the boundary between the bulk plasma region and the sheath region are transported to the sample stage, becomes shorter. Scattering probability due to collisions with volatile particles becomes small.

When the gas pressure P is increased, the mean free path λ of the ions is shortened according to the equation (2). From this point of view, the scattering probability due to collision of the ions with neutral particles increases.

When the gas pressure P is low and the frequency f is high and P / f is small, the scattering probability due to collision of ions with neutral particles in the sheath region can be reduced from the equation (3). For this reason, when the gas pressure P is low and the frequency f is high, the energy attenuation of the ions is suppressed, the directionality of the ions is aligned, and the ions are made to enter the sample substantially perpendicularly.
Further, it is possible to suppress the attenuation of the ion flux density reaching the sample, and improve the etching throughput and realize sufficient etching anisotropy.

When the gas pressure P and the frequency f are lowered,
From equation (1), the sheath width d becomes long, and from this point of view, the scattering probability due to collision of ions with neutral particles becomes large.

When the gas pressure P is lowered, the mean free path λ of the ions becomes longer according to the equation (2), so from this viewpoint, the scattering probability due to collision of the ions with neutral particles becomes small.

When the gas pressure P is high and the frequency f is low and P / f is large, the scattering probability due to collision of ions with neutral particles in the sheath region can be increased from equation (3). Therefore, when the gas pressure P is high and the frequency f is low, the energy of the ions is attenuated, the directionality of the ions is made slightly disordered, and the ions are made incident on the sample.
Further, the ion flux density reaching the sample can be attenuated, and thereby the etching ability can be relaxed.

(E) Next, the above-mentioned relationship centering on the external operating parameter called the sample stage temperature will be described.

The adsorption rate of the reaction product radicals existing in the chamber and the side wall protection radicals generated from the side wall protection additive gas on the side wall of the line pattern generally increases as the temperature of the side wall of the line pattern (that is, the sample temperature) increases. Get smaller.

On the side wall of the isolated line pattern, the expected solid angle for the side wall protective radical flux incident isotropically from above the sample is approximately π / 2, which is sufficiently large, so that the side wall protective radical is isolated line pattern. Reach the bottom sidewall well.

However, in the side wall of the internal line pattern, the expected solid angle is small, and particularly when the pattern aspect ratio is large, the expected solid angle is considerably small, so that the side wall protecting radical reaches the side wall below the inner line pattern. Must repeat adsorption and re-emission on the sidewalls of the inner line pattern several times. In this case, if the temperature of the line pattern side wall (that is, the sample temperature) is low, the adsorption rate becomes large, and most of the side wall protecting radicals adhere to the resist mask side wall on the upper side of the internal line pattern, and the lower portion of the internal line pattern The side wall protecting radicals do not reach the side walls sufficiently. On the other hand, when the temperature of the side wall of the line pattern is high, the adsorption rate becomes small, and the side wall protecting radical reaches the side wall below the inner line pattern sufficiently.

A detector that evaluates the ratio of side wall protective radicals that play a role of side wall protection in plasma, a detector that evaluates the ion flux and energy distribution of ions that play a role of etching the deposited side wall protective film, and The plurality of external control parameters described above can be optimized using signals output from a sheath width detector or the like that evaluates the angular distribution.

Further, the completion of the main etching is judged by using the signal of the etching end point detector installed in the plasma generation chamber, etching is performed under the main etching conditions until the main etching is completed, and thereafter, the over etching conditions are used. It is programmed so that automatic two-step etching can be performed, as in the case of etching.

The solving means specifically taken by the present invention will be described below.

According to the first aspect of the present invention, the means for solving the problems is to etch a sample to be etched, which is placed on the sample table and has a resist pattern formed on the surface thereof in a vacuum chamber having a sample table at the bottom. A raw material gas comprising an etching gas for etching and a sidewall protecting gas for generating sidewall protecting radicals for protecting the sidewall of the line pattern formed by etching the sample to be etched A dry etching method in which ions are generated and high-frequency power is applied to the sample stage to form a self-DC bias to guide the ions to the sample stage, thereby etching the sample to be etched. , A line pattern group consisting of a plurality of the line patterns formed close to each other The line width of the first line pattern formed of the line pattern located on the inner side is formed of the line pattern located on the outermost side of the line pattern group or the line pattern formed isolated from the line pattern group. The line width is narrower than the line width of the second line pattern, and the first and second line patterns are formed.
When the line width of the line pattern is thicker than the line width of the resist pattern, the etching amount on the sidewalls of the first and second line patterns increases and the etching amount on the sidewalls of the first line pattern increases. The gas pressure of the raw material gas introduced into the vacuum chamber, the discharge amount of the gas discharged from the vacuum chamber, and the frequency of the high frequency power so that the etching amount relative to the sidewall of the second line pattern is relatively reduced. And a parameter control step of changing at least one parameter of a parameter group consisting of the power of the high-frequency power, the ratio of the sidewall protecting gas in the raw material gas, and the temperature of the sample stage. Is.

According to the structure of claim 1, in the parameter control step, the etching amount for the first and second line patterns is increased, and the etching amount for the first line pattern is relative to the etching amount for the second line pattern. The line width of the first line pattern approaches the line width of the second line pattern, and the line widths of the first and second line patterns approach the line width of the resist pattern in order to control so that the line width of the first line pattern decreases. .

According to a second aspect of the present invention, in addition to the configuration of the first aspect, the parameter control step includes a step of increasing the amount of gas discharged from the vacuum chamber. .

According to the second aspect of the present invention, since the amount of gas discharged from the vacuum chamber is increased, the number of side wall protective radicals in the vacuum chamber is reduced, and particularly the side wall protective film deposited on the side wall of the second line pattern. Is decreased, the etching effect of the obliquely incident ions on the side wall of the second line pattern is relatively increased.

According to a third aspect of the present invention, in addition to the configuration of the first aspect, the parameter control step includes a step of reducing the proportion of the sidewall protecting gas in the raw material gas. is there.

According to the third aspect of the present invention, since the ratio of the side wall protecting gas to the source gas is reduced, the side wall protecting radicals in the vacuum chamber are reduced, and particularly the side wall protecting deposited on the side wall of the second line pattern. Since the amount of the film is reduced, the etching effect of the obliquely incident ions on the side wall of the second line pattern is relatively increased.

According to a fourth aspect of the present invention, in the structure of the first aspect, the parameter control step increases the gas pressure of the raw material gas introduced into the vacuum chamber, and increases the high-frequency power. And a configuration having "and" is added.

According to the structure of claim 4, since the gas pressure of the raw material gas introduced into the vacuum chamber is increased and the electric power of the high frequency power is increased, the decrease of the ion energy with the increase of the gas pressure increases the electric power of the high frequency power. Is compensated for by the obliquely incident ions, and the etching effect on the side wall protective film by the obliquely incident ions is increased, and the spread of the angular distribution of the incident ions is increased, and the etching effect on the side wall protective film of the first line pattern is relatively reduced.

According to a fifth aspect of the present invention, in the structure of the first aspect, the parameter control step includes a step of increasing the gas pressure of the raw material gas introduced into the vacuum chamber, and the discharge of gas discharged from the vacuum chamber. And the step of increasing the amount.

According to the structure of claim 5, since the gas pressure of the source gas introduced into the vacuum chamber is increased and the discharge amount of the gas discharged from the vacuum chamber is increased, the side wall protecting radicals in the vacuum chamber are reduced. As the amount of the sidewall protection film deposited on the sidewalls of the line pattern decreases, the spread of the angle distribution of the incident ions increases, and the etching effect on the sidewall protection film of the first line pattern relatively decreases.

According to a sixth aspect of the present invention, in addition to the configuration of the first aspect, the parameter control step includes a step of increasing the temperature of the sample stage.

According to the structure of claim 6, since the temperature of the sample stage is raised, the side wall protecting radicals particularly deposited on the side wall of the second line pattern are reduced.

The invention of claim 7 is directed to the same dry etching method as that of the invention of claim 1, wherein the line patterns located inside of a line pattern group consisting of a plurality of the line patterns formed close to each other. The line width of the first line pattern formed of the line pattern is the outermost line pattern in the line pattern group, or the line width of the second line pattern formed of the line pattern isolated from the line pattern group. When the line width of the first and second line patterns is smaller than that of the resist pattern, the etching amount of the sidewalls of the first and second line patterns is reduced. The etching amount on the sidewall of the first line pattern is equal to the second line pattern. Of the source gas introduced into the vacuum chamber, the amount of gas exhausted from the vacuum chamber, the frequency of the high-frequency power, A parameter control step of changing at least one parameter of a parameter group consisting of electric power, the ratio of the sidewall protecting gas in the raw material gas, and the temperature of the sample stage is provided.

According to the structure of claim 7, in the parameter control step, the etching amount for the first and second line patterns is reduced, and the etching amount for the first line pattern is relative to the etching amount for the second line pattern. The line width of the first line pattern approaches the line width of the second line pattern, and the line widths of the first and second line patterns approach the line width of the resist pattern in order to control so that the line width of the first line pattern decreases. Therefore, the line widths of the first and second line patterns are optimized.

According to an eighth aspect of the present invention, in addition to the configuration of the seventh aspect, the parameter control step includes a step of reducing the gas pressure of the raw material gas introduced into the vacuum chamber. Is.

According to the structure of claim 8, since the gas pressure of the source gas introduced into the vacuum chamber is reduced, the obliquely incident ions are reduced and the etching effect of the obliquely incident ions on the side wall protective film is reduced. And the amount of dimensional reduction in the second line pattern is reduced.

According to a ninth aspect of the invention, in addition to the configuration of the seventh aspect, the parameter control step includes a step of reducing the amount of gas discharged from the vacuum chamber. .

According to the ninth aspect, since the amount of gas discharged from the vacuum chamber is reduced, the number of side wall protective radicals in the vacuum chamber increases and the amount of side wall protective film deposited on the side wall of the line pattern increases. At the same time, the reduction of the etching effect on the side wall protective film due to the obliquely incident ions appears remarkably in the second line pattern.

According to a tenth aspect of the present invention, there is provided the structure of the seventh aspect.
The parameter control step adds a configuration including a step of raising the temperature of the sample stage.

According to the structure of the tenth aspect, since the temperature of the sample stage is increased, the side wall protecting radicals deposited on the side wall of the line pattern are reduced.

The invention of claim 11 is directed to the same dry etching method as that of the invention of claim 1, and the line patterns located inside of a line pattern group consisting of a plurality of the line patterns formed close to each other. The line width of the first line pattern formed of the line pattern is the outermost line pattern in the line pattern group, or the line width of the second line pattern formed of the line pattern isolated from the line pattern group. When the line width of the first and second line patterns becomes thicker than the line width of the resist pattern, the etching amount of the sidewalls of the first and second line patterns increases. The etching amount on the side wall of the first line pattern is equal to that of the second line pattern. Gas pressure of the source gas introduced into the vacuum chamber, the amount of gas exhausted from the vacuum chamber, the frequency of the high frequency power, the high frequency power so that the etching amount relative to the side wall of the furnace increases. And a parameter control step of changing at least one parameter of a parameter group consisting of the power, the ratio of the side wall protection gas in the raw material gas, and the temperature of the sample stage.

According to the eleventh aspect, in the parameter control step, the etching amount for the first and second line patterns is increased, and the etching amount for the first line pattern is relative to the etching amount for the second line pattern. The line width of the first line pattern approaches the line width of the second line pattern, and the line widths of the first and second line patterns approach the line width of the resist pattern in order to control so that the line width of the first line pattern increases. Therefore, the line widths of the first and second line patterns are optimized.

According to a twelfth aspect of the present invention, in the configuration of the eleventh aspect, the parameter control step includes a step of increasing the power of the high frequency power and a step of increasing the amount of gas discharged from the vacuum chamber. The configuration of having is added.

According to the twelfth aspect, since the high frequency power is increased and the amount of gas discharged from the vacuum chamber is increased, the side wall protecting radicals in the vacuum chamber are reduced and deposited on the side wall of the line pattern. Since the amount of the side wall protective film is reduced, the size reduction amount of the first and second line patterns is increased, and the spread of the angular distribution of incident ions is reduced, so that the etching effect on the side wall protective film of the first line pattern is reduced. Is relatively increased, the dimension reduction amount of the first line pattern is greater than the dimension reduction amount of the second line pattern.

According to a thirteenth aspect of the present invention, in addition to the configuration of the eleventh aspect, the parameter control step includes a step of reducing the gas pressure of the raw material gas introduced into the vacuum chamber. Is.

According to the thirteenth aspect, since the gas pressure of the source gas introduced into the vacuum chamber is reduced, the obliquely incident ions are reduced, and the etching effect on the side wall protective film by the obliquely incident ions is reduced by the second line. Remarkably appears in the pattern.

According to a fourteenth aspect of the present invention, in addition to the configuration of the thirteenth aspect, the parameter control step further includes a step of increasing a discharge amount of gas discharged from the vacuum chamber. is there.

According to the structure of claim 14, since the gas pressure of the raw material gas introduced into the vacuum chamber is reduced and the discharge amount of the gas discharged from the vacuum chamber is increased, the side wall protective radicals in the vacuum chamber are reduced. As the amount of the sidewall protection film deposited on the sidewalls of the line pattern decreases, the spread of the angle distribution of the incident ions becomes small, and the etching effect on the sidewall protection film of the first line pattern relatively increases.

According to a fifteenth aspect of the present invention, in addition to the configuration of the eleventh aspect, the parameter control step includes a step of increasing the frequency of the high frequency power.

According to the fifteenth aspect, since the frequency of the high frequency power is increased, the sheath width is reduced, the ratio of obliquely incident ions is decreased, and the etching effect on the side wall protective film by the obliquely incident ions is secondly reduced. Remarkably appears in the line pattern.

According to a sixteenth aspect of the present invention, in addition to the configuration of the eleventh aspect, the parameter control step includes a step of increasing the temperature of the sample stage.

According to the sixteenth aspect, since the temperature of the sample stage is raised, the side wall protecting radicals deposited on the side wall of the line pattern are reduced.

The invention of claim 17 is directed to the same dry etching method as that of the invention of claim 1, and the line patterns located inside of a line pattern group consisting of a plurality of the line patterns formed close to each other. The line width of the first line pattern formed of the line pattern is the outermost line pattern in the line pattern group, or the line width of the second line pattern formed of the line pattern isolated from the line pattern group. When the line width of the first and second line patterns is smaller than the line width of the resist pattern, the etching amount on the sidewalls of the first and second line patterns is reduced and The etching amount on the side wall of the first line pattern is equal to that of the second line pattern. Gas pressure of the source gas introduced into the vacuum chamber, the amount of gas exhausted from the vacuum chamber, the frequency of the high frequency power, the high frequency power so that the etching amount relative to the side wall of the furnace increases. And a parameter control step of changing at least one parameter of a parameter group consisting of the power, the ratio of the side wall protection gas in the raw material gas, and the temperature of the sample stage.

According to the seventeenth aspect, in the parameter control step, the etching amount for the first and second line patterns is reduced, and the etching amount for the first line pattern is relative to the etching amount for the second line pattern. The line width of the first line pattern approaches the line width of the second line pattern, and the line widths of the first and second line patterns approach the line width of the resist pattern in order to control so that the line width of the first line pattern increases. .

According to an eighteenth aspect of the present invention, in addition to the configuration of the seventeenth aspect, the parameter control step includes a step of reducing the amount of gas discharged from the vacuum chamber. .

According to the eighteenth aspect of the present invention, since the amount of gas discharged from the vacuum chamber is reduced, the amount of side wall protective radicals in the vacuum chamber increases and the amount of the side wall protective film deposited on the side wall of the line pattern is particularly reduced. The line width of the second line pattern relatively increases, so that the line widths of the first and second line patterns approach the line width of the resist pattern even when the etching effect of the obliquely incident ions on the sidewall protection film is taken into consideration.

According to a nineteenth aspect of the present invention, in addition to the configuration of the seventeenth aspect, the parameter control step includes a step of increasing the proportion of the sidewall protecting gas in the raw material gas. Is.

According to the nineteenth aspect of the present invention, since the ratio of the side wall protective gas to the source gas is increased, the side wall protective radicals in the vacuum chamber increase and the amount of the side wall protective film deposited on the side wall of the line pattern is particularly increased. Since it increases relatively in the second line pattern, even if the etching effect on the side wall protection film by obliquely incident ions is considered,
The line widths of the first and second line patterns approach the line width of the resist pattern.

The invention of claim 20 is the addition of the structure of claim 17, wherein the parameter control step includes a step of increasing the frequency of the high frequency power.

According to the twentieth aspect, since the frequency of the high frequency power is increased, the sheath width is reduced, the ratio of obliquely incident ions is reduced, and the second effect is to reduce the etching effect of the obliquely incident ions on the side wall protective film. Remarkably appears in the line pattern.

According to a twenty-first aspect of the invention, the parameter control step further includes a step of reducing the gas pressure of the raw material gas introduced into the vacuum chamber, in addition to the configuration of the twentieth aspect. It is a thing.

According to the twenty-first aspect, since the frequency of the high frequency power is increased and the gas pressure of the raw material gas introduced into the vacuum chamber is reduced, the obliquely incident ions are greatly reduced and the side wall protection film by the obliquely incident ions is reduced. While the etching effect is reduced, the reduction of the etching effect on the side wall protection film due to the obliquely incident ions remarkably appears in the second line pattern.

The invention of claim 22 is the addition of the structure of claim 17, wherein the parameter control step includes a step of lowering the temperature of the sample stage.

According to the structure of claim 22, since the temperature of the sample stage is lowered, the amount of the side wall protective film deposited on the side wall of the line pattern relatively increases, especially in the second line pattern. Even considering the etching effect on the sidewall protective film, the line widths of the first and second line patterns approach the line width of the resist pattern.

According to a twenty-third aspect of the present invention, in addition to the configuration of the seventeenth aspect, the parameter control step includes a step of reducing the gas pressure of the raw material gas introduced into the vacuum chamber. Is.

According to the twenty-third aspect, since the gas pressure of the raw material gas introduced into the vacuum chamber is reduced, the obliquely incident ions are reduced, the etching effect on the side wall protective film by the obliquely incident ions is reduced, and the oblique incidence is performed. The reduction of the etching effect on the side wall protection film due to the ions remarkably appears in the second line pattern.

[0099]

BEST MODE FOR CARRYING OUT THE INVENTION A parallel plate type reactive ion dry etching apparatus which is an example of the dry etching apparatus used in the present invention will be described below.

FIG. 1 shows a schematic configuration of a parallel plate type reactive ion dry etching apparatus. As shown in FIG. 1, a gas controller 1 is provided in a metal chamber 11.
A reactive gas is introduced through the inside of the chamber 11, and the inside of the chamber 11 is controlled to an appropriate pressure by the exhaust system 13.

An anode (anode) 14 is provided on the upper portion of the chamber 11, and a sample table 15 serving as a cathode (cathode) is provided on the lower portion. A high-frequency power supply source 17 for supplying high-frequency power is connected to the sample stage 15 via an impedance matching circuit 16, and high-frequency discharge can be generated between the sample stage 15 and the anode electrode 14. The frequency of the high frequency supplied from the high frequency power supply source 17 can be changed by the frequency control circuit 21.

The energy distribution of ions and the width of the sheath region near the sample stage 15 can be judged by the plasma parameter detector 26, and the etching end point can be judged by the etching end point detector 20 using the spectral method. . Further, the gas controller 12 and the exhaust system 13 are controlled by a signal from the etching end point detector 20, and the gas pressure and the exhaust amount in the chamber 11 are appropriately controlled. Further, the frequency of the high frequency power source 17 is controlled via the frequency control circuit 21 by the signal from the etching end point detector 20.

The temperature of the sample stage 15 can be adjusted by controlling the heater 23 via the temperature control circuit 24. In addition, the external parameter control device 22 controls the signal from the plasma parameter detector 26 and the signal from the etching end point detector 20, or the value of these signals and the external parameters such as frequency, gas pressure, high frequency power power, and sample stage temperature. It is possible to control the gas controller 12, the exhaust system 13, the frequency control circuit 21, and the temperature control circuit 24 based on a combination of the above, or a combination of these signals and a pre-programmed processing flow.

In the plasma generation region surrounded by the anode electrode 14 and the sample table 15, the anode electrode 14
The region except the sheath region near the sample table 15 is generally called a bulk plasma region.

FIG. 2 shows that in the parallel plate type reactive ion etching apparatus shown in FIG. 1, ions starting from the boundary between the bulk plasma region and the sheath region are accelerated toward the sample stage 15 between the sheath widths d. While being transported to the sample table 15 while being transported, the ions are scattered by collision with neutral particles, the directionality of the ions becomes disordered, and the energy of the ions is attenuated.

For the same frequency, collision scattering between ions and neutral particles is smaller in the high vacuum where the gas pressure is low than in the low vacuum where the gas pressure is high.
Due to the high energy, the high ion flux enters the sample stage 15 more vertically. On the other hand, collisions and scattering of ions and neutral particles occur more frequently in the case of low vacuum with high gas pressure than in the case of high vacuum with low gas pressure, so the ion flux density decreases and the energy of ions decreases. Decreased,
The incident angle distribution of the ions on the sample table 15 is widened.

For the same gas pressure, ions and neutral particles collide at a relatively high frequency where the sheath width d becomes shorter than at a relatively low frequency where the sheath width d becomes longer. Since the scattering is reduced, the high ion flux enters the sample stage 15 more vertically due to the high energy. On the other hand, when the sheath width d is relatively low and the sheath frequency d is relatively low, the collision scattering between the ions and the neutral particles occurs more frequently than the case where the sheath width d is relatively high and the ion flux density is high. Becomes smaller, the ion energy decreases, and the incident angle distribution of the ions on the sample table 15 expands.

3 to 10 show, in the parallel plate type reactive ion etching apparatus shown in FIG. 1, starting from the boundary between the bulk plasma region and the sheath region, being accelerated by the electric field in the sheath region, and being neutral particles. Ion angular distribution f (θ, E _i ) integrated as a function of the energy E _i of chlorine ions Cl ⁺ and the angle θ reaching the wafer surface on the sample stage 15 while colliding and in the energy direction. The solid line indicates the ion energy distribution h (E _i ) obtained by integrating g (θ) and the ion angular energy distribution f in the angular direction. The ion angle θ is an angle measured from a direction perpendicular to the wafer surface. 3 to 10, (a) shows the ion angle distribution, and (b) shows the ion energy distribution. However, sheath width L _sh = 1 cm, voltage between sheaths = 200 V,
The gas pressure P is 0.1 to 10 Pa.

As can be seen from FIGS. 3 to 10, when the gas pressure is 0.1 Pa or 0.2 Pa, the angular distribution is concentrated near 0 degrees, that is, in the vertically incident state, and scattering other than 0 degrees is observed. The component is very small, the energy distribution is mostly concentrated near the sheath-to-sheath voltage of 200 V, and the low energy component decelerated by scattering is very small.

When the gas pressure is increased from 0.5 Pa to 2 Pa, the scattered component sharply increases. The peak near 200 V corresponding to the inter-sheath voltage gradually decreases, and conversely, the low energy component decelerated by scattering relatively increases.

When the gas pressure is increased from 5 Pa to 10 Pa, the scattering component certainly increases in the ion angle distribution, but the scattering component is not so different from the case where the gas pressure is 2 Pa. On the other hand, in the ion energy distribution, the peak near 200 V disappears and the center of the energy distribution moves to the low energy side.

In FIGS. 3 to 10, the angular distribution curve g ^* (θ) shown by the broken line represents the weight of the reaction yield y (the number of Si atoms sputtered / the number of incident ions) due to the energy-dependent Cl ⁺ beam. It is obtained by multiplying the ion angular energy distribution f and integrating it in the energy direction. That is, the angle distribution curve has different effects on etching between ions having an energy of 50 eV and ions having an energy of 100 eV even if the ions are incident on the wafer at the same angle θ. The aim is to draw an effective angle distribution curve that incorporates this effect. Here, the incident energy E _i of the beam is e
It is a unit of V. Specifically, the incident energy E _i (e
V) corresponds to the inter-sheath voltage Vs (V).

As can be seen from FIGS. 3 to 10, the ion E having a large scattering angle component generally has a reduced energy E _i , and therefore the weight y (E _i ) of the reaction yield also becomes small. As a result, the spread of the ion angle distribution shown by the broken line in consideration of the weight of the reaction yield y is smaller than the ion angle distribution g obtained by simply integrating the energy.

FIG. 11 shows the standard deviation σ when the normal distribution representing the spread of the scattering angles in the ion angle distributions shown in FIGS. 3 to 10 is assumed with the gas pressure as the horizontal axis. The standard deviation σ of the ion angle distribution increases as the gas pressure increases. That is, the ion angle distribution is
As the gas pressure increases, it gradually expands from vertical incidence. When the gas pressure becomes 2 Pa or less,
The standard deviation σ decreases almost logarithmically with decreasing gas pressure. On the other hand, when the gas pressure is 2 Pa or more,
Even if the gas pressure increases, the standard deviation σ is almost saturated, and the increasing rate is gentle. These changes correspond to the changes in the mean free path λ of the ions which are simultaneously shown. Standard deviation σ
Can be read to be approximately 10 degrees when the gas pressure is 0.1 Pa, approximately 24 degrees when the gas pressure is 1 Pa, and approximately 27 degrees when the gas pressure is 10 Pa.

It can be understood that when the plasma internal parameters other than the gas pressure are made constant, the spread of the ion angle distribution can be controlled to some extent by changing the gas pressure around 1 Pa.

The standard deviation σ of the ion angle distribution was evaluated, but as can be understood from FIGS. 3 to 10, the ion angle distribution is different from the so-called normal distribution curve. In this sense, in order to grasp the features of the ion angle distribution more intuitively, consider a certain finite scattering angle width Δ (degrees),
The peaking ratio R, which is the ratio of the number of ions having a scattering angle smaller than the scattering angle width Δ to the number of all ions
Evaluate (Δ).

FIG. 12 shows the peaking ratio R (Δ) in the ion angle distributions shown in FIGS. 3 to 10 with the gas pressure as the horizontal axis. The peaking ratio R (Δ) increases as a result of a decrease in ion scattering with a decrease in gas pressure. As with the change in standard deviation σ, especially gas pressure is 2P
In the case of a or less, the peaking ratio R (Δ) increases remarkably as the gas pressure decreases. On the other hand, when the gas pressure is 2 Pa or more, the peaking ratio R (Δ) is almost saturated, and the peaking ratio R (Δ) decreases gradually with the increase of the gas pressure. In FIG. 12, a black symbol is a case where the weight of the reaction yield y is taken into consideration, and a white symbol is a case where the weight of the reaction yield y is not taken into consideration. Hereinafter, the case where the weight of the reaction yield y is taken into consideration will be examined.

Consider the case where the gas pressure is 1 Pa. For example, the relative proportion of ions within the scattering angle width Δ = ± 1 degree, that is, the peaking ratio R (Δ = 1 °) at which the ions reach the wafer with almost no collision with neutral particles.
Is about 17%, and the relative peaking ratio R (Δ = 5 °) of the ions within the scattering angle width Δ = ± 5 degrees is about 30%. This means that when the gas pressure is 0.1 Pa, the relative peaking ratio R (1 °) of the ions within the scattering angle width Δ = ± 1 degree is about 82%, and the scattering angle width Δ = The relative peaking ratio R (5 °) of the ions in the range of ± 5 degrees sharply increases to about 86%.

When the gas pressure P is 0.2 Pa or less,
The ion angle distribution is in a state of almost vertical incidence, and the scattered components other than 0 degree are very small. Gas pressure from 0.5 Pa to 2
When the pressure is increased to Pa, the scattered component increases significantly. When the gas pressure P is increased from 5 Pa to 10 Pa, the scattered component is certainly increased, but there is not much difference as compared with the case of P = 2 Pa.

When the gas pressure P is 0.2 Pa or less,
The ion energy distribution is equivalent to the inter-sheath voltage of 200
Almost concentrated in the vicinity of V, there are very few low energy components decelerated by scattering. When the gas pressure is increased from 0.5 Pa to 2 Pa, the voltage corresponding to the sheath-to-sheath voltage becomes 20.
The peak near 0 V gradually decreases, and conversely, the low energy component decelerated by scattering relatively increases.
If the gas pressure is increased from 5 Pa to 10 Pa,
The peak near 0 V disappears, and the center of the energy distribution gradually moves toward the low energy side.

In general, the incident ion having a large scattering angle component also attenuates the energy E _i, so that the reaction yield y
Also becomes smaller. As a result, the spread of the ion angle distribution considering the weight of the reaction yield y is smaller than that of the ion angle distribution simply integrated in the energy direction. When considering the effect on etching, the ion angle distribution in which the weight of the reaction yield y is taken into consideration is more significant.

The standard deviation σ of the angular distribution increases as the gas pressure increases. That is, the angular distribution gradually expands from the vertically incident state as the gas pressure increases. When the gas pressure is 2 Pa or less, the standard deviation σ increases almost logarithmically as the gas pressure increases. On the other hand, when the gas pressure is 2 Pa or more, the standard deviation σ is almost saturated, and the increasing rate is moderate. These changes correspond to changes in the mean free path λ of the ions. The standard deviation σ is about 10 degrees when the gas pressure is 0.1 Pa, and is about 24 degrees when the gas pressure is 1 Pa,
It can be read that it is about 27 degrees when the gas pressure is 10 Pa.

Summarizing the above results, the sheath width is 1c.
When m is set, the frequency of collision with the neutral particles of ions is sharply reduced below the point where the gas pressure becomes 2 Pa, and the ion energy distribution h becomes a value corresponding to the voltage between the sheaths. The ion angle distribution g is concentrated and is in a state of almost vertical incidence in which scattering components other than 0 degree are very small.
On the other hand, when the gas pressure is 2 Pa or more, the decelerated low energy component relatively increases and the vertically incident component relatively decreases.

It should be noted that the spread of the ion angle distribution can be controlled to some extent by changing the gas pressure before and after 1 Pa under the condition that the parameters inside the plasma other than the gas pressure are constant. This suggests a possibility for flexible etching shape control in this gas pressure region.

The results of FIGS. 3 to 12 shown above have different values quantitatively if the voltage applied to the electrodes, the boundary conditions such as walls and the gas species used are changed. However, qualitatively, the above-mentioned tendencies are almost reproduced.

Next, the frequency f of the high frequency power and the gas pressure P
The above-mentioned relationship centered on the combination of the two external operating parameters will be described.

That is, the cathode-side sheath width d formed in the vicinity of the sample table 15 is d = K ₁ / (P ^m · f ⁿ ) (where K ₁ is a constant), where m is positive. Is a real number greater than approximately 1/3 and less than approximately 1/2, and n is a positive real number greater than approximately 1/2 and less than approximately 1. This is K. Harafuji, A. Yamano and M. Kubota: Jpn. J. App.
l. Phys. vol.33 (1994) p2212, and N. Mutsukura, KK
obayashi and Y.Machi: J. Appl. Phys.vol.68 (1990)
It has already been explained on p.2657.

Further, the mean free path λ of ions mainly derived from elastic collision scattering and charge exchange scattering between ions and neutral particles is inversely proportional to the gas pressure P and can be expressed as follows. That is, λ = K ₂ / P (where K
₂ is a constant. ) Further, it is proportional to the probability that ions that have left the boundary between the bulk plasma region and the sheath region are scattered by collision with neutral particles in the sheath region while being transported to the sample stage 15 that is the cathode electrode. The quantity η is η = d / λ. Therefore, by substituting the expressions of d and λ into the expression of η, η = d / λ = (K ₁ / (P ^m · f ⁿ )) × (P / K ₂ ) = (K ₁ / K ₂ ) × (P ^1-m / f ⁿ )-(K ₁ / K ₂ ) × (P / f) ^1/2 .

That is, when the gas pressure P and the frequency f are increased, the sheath width d, which is the distance traveled while the ions starting from the boundary between the bulk plasma region and the sheath region are transported onto the sample stage 15, becomes shorter. . From this point of view, the scattering probability due to collision of ions with neutral particles is small.

When the gas pressure P is increased, the mean free path λ of ions is shortened, so that the scattering probability due to collision of ions with neutral particles increases. Therefore, by lowering the gas pressure P and increasing the frequency f to reduce P / f, it is possible to reduce the scattering probability due to collision of ions with neutral particles in the sheath region. As a result, the energy attenuation of the ions can be suppressed, and the ion flux density reaching the sample table 15 can be suppressed by making the ions have the same directionality and being incident substantially perpendicularly to the sample table 15. It is possible to improve the etching throughput and achieve sufficient etching anisotropy.

When the gas pressure P and the frequency f are lowered, the sheath width d becomes longer, so that the probability of scattering due to collision of ions with neutral particles increases. Further, when the gas pressure P is lowered, the mean free path λ of ions becomes longer. For this reason, the scattering probability due to collision of ions with neutral particles becomes small.
Therefore, by increasing the gas pressure P and decreasing the frequency f, P
By increasing / f, it is possible to increase the scattering probability due to collision of ions with neutral particles in the sheath region. As a result, the energy of the ions is attenuated, and the directionality of the ions is slightly disturbed, so that the sample stage 15
Since the ion flux density reaching the sample stage 15 can be attenuated by making the light incident on the substrate, the etching ability can be relaxed.

The amount of the side wall protective film deposited on the side wall of the pattern due to the reaction product radicals in the chamber is larger in the isolated line pattern (or the outer line pattern) than in the inner line pattern, regardless of the gas pressure in the chamber. There are also many. Further, the amount of change in the sidewall protective film deposition amount with respect to the change in the ratio of reaction product radicals in the chamber is larger in the side wall of the isolated line pattern than in the side wall of the internal line pattern.

In the medium vacuum region, the shaving effect of the side wall protective film by obliquely incident ions is larger in the isolated line pattern than in the inner line pattern side wall due to the ion collimation effect. The change in the ion incident angle appears as a larger change in the etching amount of the side wall protective film on the side wall of the isolated line pattern. That is, with respect to the ratio of reaction product radicals and changes in the ion incident angle,
The isolated line pattern is relatively sensitive and the inner line pattern is relatively insensitive.

In the high vacuum region, the ions are incident on the wafer surface with a relatively small scattering angle. Due to this "quasi-parallel beam effect", the ion flux incident on the side wall of the isolated line pattern is larger than the ion flux incident on the side wall of the internal line pattern, but the difference is not as remarkable as in the medium vacuum region. In particular, the difference in etching ability of the deposition protection film between the isolated line pattern and the internal line pattern is not so different as in the medium vacuum region. Further, in the high vacuum region, the angle of incident ions with respect to the pattern side wall is relatively small, and the proportion of ions incident on the pattern side wall at a large angle is small. Therefore, on the side wall of the isolated line pattern, the ability to etch the deposition protection film is smaller in the high vacuum than in the medium vacuum.

According to the above description, the mechanism is different regardless of whether it is a medium vacuum or a high vacuum, but the profile of the pattern is made vertical and the dimensional change between the line pattern inside the line and space and the isolated line pattern is different. In order to reduce the amount of exhaust gas, controlling the amount of exhaust gas to control the proportion of reaction product radicals in the chamber and controlling the amount of deposition of the sidewall protective film is one effective measure. Understandable.

Next, the above-mentioned relationship centering on the external operating parameter of the temperature of the sample table 15 will be described.

The adsorption rate of radicals, such as reaction product radicals existing in the chamber 11 and radicals generated from the sidewall protection additive gas, which play a role of sidewall protection on the pattern sidewall is generally the temperature of the pattern sidewall (ie , Sample temperature) becomes smaller. In the case of the isolated line pattern side wall, the estimated solid angle for the side wall protective radical flux that is incident isotropically from above the wafer is about π / 2, which is sufficiently large. The side wall protecting radicals fully reach the side walls of. However, in the case of the inner line pattern side wall, the expected solid angle becomes considerably small,
Particularly, when the pattern aspect ratio (the value obtained by dividing the line height of the line-and-space pattern by the space width) is large, the expected solid angle becomes considerably small. As a result, in order for the side wall protective radicals to reach the side wall of the sample to be etched below the inner line pattern, adsorption and re-emission must be repeated several times on the side wall of the inner line pattern. At this time, if the temperature of the pattern side wall (that is, the sample temperature) is low, the adsorption rate becomes large. Therefore, most of the radicals that play a role of protecting the side wall adhere to the side wall of the resist mask above the internal line pattern and protect the side wall. Radicals do not fully reach the sidewall of the sample to be etched below the internal line pattern. On the other hand, when the temperature of the pattern side wall (that is, the sample temperature) is high, the adsorption rate becomes small, so that the side wall protecting radicals sufficiently reach the side wall of the sample to be etched below the internal line pattern.

FIG. 13 shows the movement of the side wall protecting radicals 40 incident isotropically from above the line-and-space pattern in an external line pattern in which an environment equivalent to the isolated line pattern is established, and in the inside. This is explained in the form of comparison with the line pattern. In FIG. 13, 30 is a photoresist pattern, 31 is a phosphorus-doped polycrystalline silicon film, 32 is a thermal oxide film, and 33 is a silicon substrate.

During the main etching, the ions or radicals react with the material to be etched to generate reaction products, and the proportion of reactive organisms increases. During overetching, the proportion of reactive products decreases. That is, since the ratio of the side wall protecting radicals in the chamber changes, the dimension and profile control methods differ.

In this example, a detector for evaluating the proportion of radicals in the plasma which plays a role of protecting the sidewall, a detector for evaluating the ion flux and ion energy distribution of ions which play the role of etching the deposited sidewall protecting film. A plurality of external control parameters described above can be optimized by using the signals of the chamber width detector and the sheath width detector that is a means for evaluating the angular distribution of ions.

Also, the chamber 11 for generating plasma
The end of the main etching is determined by using the signal of the etching end point detector 20 attached to, and etching is performed under the above-mentioned main etching conditions until the main etching is completed.
After completion of the main etching, it is programmed to automatically perform two-step etching so that the etching is performed under the above-mentioned over-etching conditions.

Next, a method will be described in which both the isolated line pattern and the internal line pattern have vertical profiles, and the dimensional shift between the two is small with respect to the resist mask, whereby the dimensional shift difference between the two is small.

FIG. 14 shows the mechanism concept of the difference in profile and dimensional change between the internal line pattern and the isolated line pattern in the line-and-space pattern in the parallel plate type reactive ion etching apparatus shown in FIG. The description is divided into a vacuum region and a high vacuum region. In FIG. 14, 34 is a side wall protective deposition film such as a reactive organism, 35 is an ion, and 36 is a film thickness reduction amount of the side wall protective deposition film such as a reaction product etched by the ion 35. Is.

In conclusion, in order to make the profile vertical and reduce the difference in dimensional change between the internal line pattern and the isolated line pattern in the line-and-space pattern, the displacement is controlled by Controlling the proportion of reaction product radicals in the chamber is one effective strategy. In the following, the medium vacuum and the high vacuum do not refer to specific pressures but are used as expressions expressing the difference in the ion angle distribution described above.

Now, regardless of the gas pressure in the chamber, the amount of the sidewall protection film deposited on the sidewall of the pattern by the sidewall protection radicals such as reaction product radicals in the chamber is
There are more isolated line patterns (or external line patterns) than internal line patterns. The reason is that, as described above, in the case of the isolated line pattern side wall, the expected solid angle for the side wall protective radical flux that is incident isotropically from above the wafer is approximately π / 2. This is because the expected solid angle becomes considerably small in the case of the inner line pattern side wall.

When the ratio of reaction product radicals in the chamber is reduced, the amount of the side wall protective film deposited on the side wall of the isolated line pattern and the amount of side wall protective film on the side wall of the internal line pattern decrease while maintaining a constant ratio. . However, the reduction rate of the sidewall protection film with respect to the reduction of the sidewall protection radical is larger in the isolated line pattern sidewall than in the internal line pattern sidewall. That is, the amount of change in the sidewall protective film on the sidewall of the isolated line pattern is larger than the amount of change in the sidewall protective film on the sidewall of the internal line pattern.

The movement of incident ions in the medium vacuum region will be examined below. In the medium vacuum region, the ions impinge on the wafer surface in a state having many relatively large scattering angle components due to collision with neutral particles in the sheath region. Such obliquely incident ions have an effect of removing the side wall protective film and making the profile vertical from the forward taper to the pattern side wall where the amount of the protective film deposited is large. The more perpendicular the ion incident angle is to the pattern sidewall, the greater the etching ability of the sidewall protective film. On the contrary, ions incident perpendicularly to the surface of the wafer have an incident angle substantially parallel to the pattern side wall, so that the side wall protective film etching ability is small. On the other hand, the obliquely incident ions scrape off the side wall protective film on the pattern side wall where the amount of the protective film deposited is small, and the protective effect is lost. As a result, the profile changes from vertical to reverse taper.

The above phenomenon is particularly remarkable as a result of direct collision of the ions with the side wall on the side wall of the isolated line pattern because the projected solid angle for obliquely incident ions is large. On the other hand, on the side wall of the internal line pattern, the expected solid angle of incidence for obliquely incident ions is considerably small, so incident ions having a scattering angle of a certain size or more do not enter the internal space of the line-and-space pattern. It will be reflected. That is, the ion angle distribution component having a large scattering angle cannot directly fly to the lower part of the side wall of the internal line pattern pattern, and only the ion angle distribution component having a relatively small scattering angle which is more perpendicular to the wafer surface. , Can be selectively incident on the surface. That is, of the ion flux having a wide ion angle distribution component, there is an effect of collimating only a component having a certain angle or less and allowing the internal pattern in the line-and-space pattern to fly to the lower side wall. This effect is the “ion collimation effect” described above.

Ions having a large scattering angle, which are somehow incident near the upper portion of the inner line pattern side wall, cannot penetrate into the lower portion of the inner line pattern side wall without being reflected several times from the side wall. During this time, the energy of the ions will decrease and the ability to etch the overcoat will decrease. In addition, the incident ion flux is naturally reduced as compared with the isolated line pattern.

The above contents will be collectively described below.

First, the change amount of the sidewall protective film deposition amount with respect to the change of the ratio of the side wall protective radicals in the chamber is larger in the isolated line pattern sidewall than in the internal line pattern sidewall. In addition, the effect of removing the sidewall protection film by obliquely incident ions is due to the ion collimation effect.
The isolated line pattern sidewall is larger than the internal pattern sidewall. Then, the change in the ion incident angle appears as a larger change in the etching amount of the side wall protective film on the side wall of the isolated line pattern. That is, the isolated line pattern is relatively sensitive to changes in the ratio of reaction product radicals and the incident angle of ions, and the internal line pattern is relatively sensitive to changes in the ratio of reaction product radicals and the changed incident angle of ions. Relatively insensitive.

Next, the movement of incident ions in the high vacuum region will be described. In the high vacuum gas pressure region, the ions are less likely to collide with the neutral particles in the sheath region, have a relatively large number of vertically incident components, and have a relatively small scattering angle as a whole. Is incident on. That is, in the high vacuum region, most of the ions are incident along the side wall of the pattern in a substantially uniform manner.

In the case of such an incident ion angle distribution, the ion flux incident on the side wall of the isolated line pattern is larger than the ion flux incident on the side wall of the internal line pattern, but the difference is about the same as in the medium vacuum region. Is not noticeable. That is, the difference in the etching ability of the deposition protection film between the isolated line pattern and the internal line pattern is not so different as in the medium vacuum region. Further, the angle of incident ions with respect to the side wall is relatively small, and the proportion of ions incident on the side wall at a large angle is small. Therefore, the ability to etch the deposition protection film is smaller than that in the medium vacuum region, and is particularly small on the side wall of the isolated line pattern. On the other hand, the size of the sidewall protective film deposition amount is larger in the isolated line pattern sidewall than in the internal line pattern sidewall. This is the "quasi-parallel beam effect" described above.

From the above description, in order to make the profile vertical and to reduce the size difference between the internal pattern and the isolated line pattern, the exhaust amount is increased and the ratio of reaction product radicals in the chamber is reduced. It can be understood that reducing the size of the sidewall protective film deposition amount is one effective measure. This is because there are few components of obliquely incident ions, and etching of the side wall protective film by ions cannot be expected so much.

FIG. 15 shows how various internal parameters change when the external control parameters such as the gas pressure P and the exhaust amount Q are changed.

When the gas pressure P is increased, the following internal parameter changes occur. First, the energy E _i of the ions incident on the wafer surface decreases. The standard deviation σ of the ion scattering angle representing the spread of the ion angle distribution g of the incident ions increases. Since the increase of the gas pressure P means the increase of the source gas, the reactive radical flux F _R and the ion flux F _i are increased as long as the input power has the ability to sufficiently ionize and excite the source gas. As a result, the proportion of side wall protective radicals such as reaction products and sputtered resist in the chamber is increased, and the side wall protective radical flux F _RP is increased.

The standard deviation σ of the ion energy E _i and the ion scattering angle is an internal parameter related to the etching ability of the deposited amount of the side wall protective film, and the decrease of the ion energy E _i causes the lowering of the etching ability and the ion scattering angle. Since the increase of the standard deviation σ of γ results in the improvement of the etching performance, it cannot be simply determined whether the increase of the gas pressure P results in the improvement or the decrease of the etching performance. On the other hand, an increase in the side wall protective radical flux F _RP obviously causes an increase in the side wall protective film deposition amount. Therefore, it cannot be simply determined whether the increase of the gas pressure P causes the increase or decrease of the sidewall protective film deposition amount. However, if the exhaust amount is increased and the ratio of the side wall protective radicals in the chamber is reduced with the gas pressure P kept constant, the etching of the side wall protective film is superior to the increase in the film deposition amount. Become. That is, the side wall protective film thickness can be controlled by controlling the exhaust amount.

Let us consider the case where both the profile of the isolated line pattern and the profile of the internal line pattern are forward tapered under a certain process condition. When there are many side wall protecting radicals in the chamber, the isolated line pattern generally has a stronger forward taper than the internal line pattern. In such a state, a method for making the profiles of the two more vertical and reducing the dimensional difference between the two will be described.

The amount Q of exhaust gas is increased and the etching of the side wall protective film makes a condition superior to the increase in the amount of film deposition. Consider a case where the sidewall protection film deposition amount is relatively large under these conditions. In the isolated line pattern, the profile that had a strong forward taper when the displacement was small changed to a vertical profile relatively greatly, and in the internal line pattern, the profile that had a forward taper when the displacement was small was compared. Slowly changes from a weak forward taper to a vertical profile.

On the other hand, consider the case where the sidewall protection film deposition amount is relatively small under the condition that the etching of the sidewall protection film is superior to the increase of the film deposition amount. In the isolated line pattern, the forward-tapered profile changes to a reverse taper when the displacement is small, and in the internal line pattern, the forward-tapered profile when the displacement is small is compared to the weak forward taper to the vertical profile. Change slowly.

That is, as described above, in the isolated line pattern side wall,
Since the amount of deposition of the side wall protective film with respect to the decrease of the ratio of the side wall protective radicals in the chamber with the increase of the exhaust amount is large, the side wall protective film etching effect is significantly exhibited by the increase of the finite standard deviation σ. Due to these effects,
It is possible to reduce the dimensional difference between the isolated line pattern and the internal line pattern when the displacement is small.

The effects of increasing the exhaust amount Q while keeping the gas pressure in the chamber constant will be summarized below. First,
Even if the exhaust amount Q is increased, the ion energy E _i and the standard deviation σ of the ion scattering angle do not change, so that the ability to etch the sidewall protective film does not change. On the other hand, since the side wall protective radicals are reduced, the film deposition amount is reduced, so that the side wall protective film etching effect is significantly exhibited by the increase of the standard deviation σ of the finite ion scattering angle.

FIG. 16 shows how various internal parameters change when the external operating parameter of the bias power W _B is changed.

When the bias power W _B is increased, the following internal parameter changes can be considered. First, the energy E _i of the ions incident on the wafer surface will increase. Furthermore, an increase in bias power W _B means an increase in ionization excitation of the source gas near the wafer surface, and therefore will result in an increase in ion flux F _i and reactive radical flux F _R. As a result, the proportion of side wall protective radicals such as reaction products and sputtered resist in the chamber is increased, and the side wall protective radical flux F _RP is increased.

The ion energy E _i , the ion flux F _i, and the standard deviation σ of the ion scattering angle are internal parameters related to the etching ability of the sidewall protective film deposition amount. The increase of the ion energy E _i and the ion flux F _i brings about the improvement of the etching ability. On the other hand, the standard deviation σ of the ion scattering angle
And the increase in the side wall protective radical flux F _{RP result in} an increase in the side wall protective film deposition amount. Therefore, the bias power W _B
It cannot be simply judged whether the increase in the amount of increase causes an increase or decrease in the amount of deposited side wall protective film. However, if the exhaust amount is increased and the ratio of the side wall protective radicals is reduced with the bias power W _B kept constant, the etching of the side wall protective film is superior to the increase in the film deposition amount. That is, the side wall protective film thickness can be controlled by controlling the exhaust amount.

Consider a case where, under certain process conditions, there are many side wall protecting radicals in the chamber, and both the profile of the isolated line pattern and the profile of the internal line pattern are forward tapered. In general, the isolated line pattern has a stronger forward taper than the internal line pattern. In such a state, let us consider a method for making the profiles of the two more vertical and reducing the dimensional difference between the two.

As shown in the table of FIG. 16, a 2 × 2 matrix of low displacement, high displacement, medium vacuum and high vacuum is considered, and the bias power W _{B is set in} each case. Consider how the profiles of the isolated line pattern and the internal line pattern change when is increased.

It is not clear whether the profile of the isolated line pattern is forward taper or reverse taper in the case of medium vacuum and low displacement, because of mutual competition. That is, when the bias power W _B is increased, the ion energy E that increases the sidewall protective film etching is increased.
_{Although i} and the ion flux F _i increase, it is considered that the increase in the amount of deposited side wall protective film is also large. Since the profile of the internal line pattern has a smaller etching ability than the isolated line pattern, the profile is from a forward taper to a weak forward taper.

When the medium vacuum is high and the displacement is high, the so-called "ion collimation effect" works. In the isolated line pattern, the side wall protective film is etched more, resulting in a change from a forward taper to an inverse taper.
On the other hand, in the internal line pattern, the etching amount of the side wall protective film is not so large, for example, from the forward taper to the weak forward taper. That is, as the bias power W _B increases, the isolated line pattern has a larger dimension reduction amount than the internal line pattern, and the pattern becomes thinner.

In the case of high vacuum and low displacement, the so-called "pseudo-parallel beam effect" works, the etching ability of the side wall protective film becomes small, and the etching ability difference between the isolated line pattern and the internal pattern becomes small. Become. However, the amount of deposited side wall protective film itself is larger in the isolated line pattern than in the internal line pattern. Therefore, if the isolated line pattern has a forward taper,
The internal line pattern has a weak forward taper. That is,
As the bias power W _B increases, the sidewall protecting radicals in the chamber increase, and the isolated line pattern becomes thicker than the internal line pattern.

The "quasi-parallel beam effect" works even when the vacuum is high and the displacement is high. The etching ability of the side wall protective film is larger than that in the case of low displacement. The difference in etching ability between the isolated line pattern and the internal pattern is small. The sidewall protective film deposition amount itself is thin due to the high exhaust amount, and the isolated line pattern has a larger deposition amount than the internal line pattern. Therefore, the bias power W _B
Assuming that the isolated line pattern changes from a forward taper to a weak forward taper due to the increase of, the internal line pattern changes from the weak forward taper to the reverse taper. Due to the increase of the bias power W _B , the internal line pattern becomes thinner than the isolated line pattern.

FIG. 17 is a summary of a method for reducing the dimensional difference between the isolated line pattern and the internal line pattern, that is, a method for reducing the aperture ratio dependence of the dimension in the case of medium vacuum.

In this case, the etching ability of the side wall protective film due to the finite effect of the standard deviation σ of the ion angle distribution is positively the “ion collimation effect” that the isolated line pattern is larger than the internal line pattern. To use. When the isolated line pattern has a stronger forward taper profile than the internal line pattern and the dimension is thick, the exhaust amount is increased or the ratio of the gas that generates the sidewall protecting radicals is decreased. As a result, the side wall protective radicals in the chamber are reduced, and the side wall protective film amount is reduced. Then, the etching effect of the sidewall protective film due to the finite effect of the standard deviation σ of the ion angle distribution is relatively increased. In this way, the isolated line pattern has a larger size reduction amount than the internal line pattern, and the intended purpose can be achieved.

Next, when the size of the isolated line pattern is smaller than that of the internal line pattern, the exhaust amount is reduced or the ratio of the gas that produces the side wall protecting radicals is increased. As a result, the number of side wall protective radicals in the chamber increases, and the amount of side wall protective film increases. With the increase of the side wall protective radical ratio, the phenomenon that "the amount of increase of the side wall protective film exceeds the amount of etching the side wall protective film due to the finite effect of the standard deviation σ of the ion angle distribution" is more likely to occur in the isolated line pattern. Since the line pattern is larger than the line pattern, the dimensional increase amount of the isolated line pattern is larger than that of the internal line pattern, and the intended purpose can be achieved.

In order to further increase the finite effect of the standard deviation σ of the ion angle distribution, the bias power W _B may be increased a little while the gas pressure is increased a little. This is for compensating for the ion energy E _i which decreases when the gas pressure is applied.

FIG. 18 is a summary of methods for reducing the aperture ratio dependence of dimensions in the case of high vacuum. In this case, the ion flux that is slightly obliquely incident has little difference in the etching ability of the sidewall protective film between the isolated line pattern and the internal line pattern, and the deposition amount of the sidewall protective film is , "Pseudo parallel beam effect" that an isolated line pattern is larger than an internal line pattern is positively used.

First, when the isolated line pattern has a stronger forward taper profile than the internal line pattern and the dimension is thicker, the exhaust amount is increased or the ratio of the gas that produces the side wall protecting radicals is decreased. As a result, the side wall protective radicals in the chamber are reduced, and the side wall protective film amount is reduced. The reduction ratio of the sidewall protection film can be achieved by utilizing the fact that the isolated line pattern is larger than the internal line pattern. On the contrary, when the isolated line pattern is thinner than the internal line pattern, the exhaust amount is decreased or the ratio of the gas that generates the side wall protecting radicals is increased.

FIG. 19 shows a method of reducing the aperture ratio dependence of the dimension in each case in this 2 × 2 matrix, considering the case of medium vacuum and high vacuum for the isolated line pattern and the internal line pattern. Is a summary of the first steps to be taken. The isolated line pattern and the internal line pattern are shown as being divided into eight types of (A)-(H) in the case of a forward taper and the case of a reverse taper under a certain process condition.

FIG. 20 shows how to deal with each of the cases (E) to (H) of FIG. 19 after that, the dimensional difference between the isolated line pattern and the internal pattern satisfies the condition. It is shown in the form of a case study.

For example, under the high vacuum condition of the case (H),
The case where the internal line pattern is inversely tapered under certain process conditions will be described. In this case, the isolated line pattern is processed as a vertical profile. As a result of performing the processing shown in case (H), the internal line pattern may become vertical and the isolated line pattern may become slightly forward tapered (state shown by (H-1)). If this condition satisfies the dimensional difference request standard, it is judged as OK and the process is terminated. If the degree of forward taper, that is, the size of the isolated line pattern is too large, try increasing the displacement Q a little (measure shown by (H-2)). If so, the inner line pattern may be slightly inversely tapered, the degree of forward taper of the isolated line pattern may be reduced, and the dimensional difference requirement may be satisfied (state indicated by (H-3)). If the dimensional difference requirement is still not satisfied, the gas pressure P is slightly increased to increase the exhaust amount, which enhances the sidewall protective film etching effect that causes the forward taper of the isolated line pattern ((H- Measures shown in 4)). By repeating such processing, the process conditions satisfying the dimensional difference requirement standard are determined.

Hereinafter, an example in which the etching process conditions are improved in the main etching operation mode by using the parallel plate type reactive ion etching apparatus according to the above-described method is schematically shown for forming a phosphorus-doped polycrystalline silicon gate. Shown in.

FIG. 21 is a pattern diagram before etching. In FIG. 21, 30 is a photoresist pattern, 31 is a phosphorus-doped polycrystalline silicon film, 32 is a thermal oxide film, and 33 is a silicon substrate. As shown in FIG. 21, the line and space pattern is 0.4 μm, and its cross section is a structure of “1.0 μm thick resist mask + 0.3 μm thick polysilicon film + underlying oxide film”, The resist mask profile has a slight forward taper. Here, etching of the resist itself is not considered. For convenience, the top surface of the resist is y = 0.

Unless otherwise specified, etching was performed under the following conditions. As the gas to be introduced into the plasma generator chamber, Cl ₂ was introduced at 40 sccm, the gas pressure P in the etching apparatus was 10 Pa, the frequency f of the high frequency power was 13.56 MHz, and the gas exhaust amount was 100.
It was performed at 0 liter / sec. In each of the following examples, the bias power, that is, the power of the high frequency power is appropriately changed and used. It is also sufficiently effective to use a gas based on another halogen-based gas such as HBr or SF ₆ .

22 to 35 show a concrete embodiment. 22 to 35, 30 is a photoresist pattern, 31 is a phosphorus-doped polycrystalline silicon film, 3
Reference numeral 4 is a sputtered photoresist or a deposited film of a reaction product of silicon and a halogen-based gas.

22 to 35, the conditions are variously changed to 2
The structure of the surface of the polycrystalline silicon film 31 which has been etched 15 times in the second step for a total of 30 seconds is shown. That is, in FIGS. 22 to 35, one surface string represents the surface shape at a certain time. The surface string has 15 layers, and each layer represents the change of the surface shape with the passage of time. Further, the height of the surface of the photoresist pattern is displayed as 0, and the unit is μm. 22 to 35, (a) shows a state before improvement, and (b) shows a state after improvement.

FIG. 22A shows the result of etching with Cl ₂ introduced at 40 sccm and a gas pressure of 10 Pa in a medium vacuum with an evacuation rate of 1000 liters / sec. As described above, the isolated line pattern has a larger amount of deposited sidewall protective film than the internal line pattern, and the pattern profile after etching has a forward taper and is thicker. FIG.22 (b) is
With the gas pressure kept constant at 10 Pa, the displacement is 20
The result was increased to 00 liters / sec. The side wall protective radicals in the chamber decreased, the amount of the protective film on the side wall of the pattern decreased, and the etching effect of the obliquely incident ions on the side wall protective film was relatively increased. As shown in the figure, the isolated line pattern has a larger size reduction amount than the internal line pattern, and the profiles of the isolated line pattern and the internal line pattern become more vertical,
The dimensional difference between the isolated line pattern and the internal line pattern is reduced.

In FIG. 23 (a), Cl ₂ is introduced at 40 sccm, sidewall protecting radical SiCl ₄ is introduced at 20 sccm, the gas pressure is set to 10 Pa and the exhaust volume is set to 1500 liters / second. As shown in the figure, the isolated line pattern has a larger deposition amount of the side wall protective film than the internal line pattern, and the pattern profile after etching has a forward taper and has a size. Are fat. FIG. 23B shows a gas pressure of 10 Pa and a displacement of 1500.
While keeping constant at l / sec, add SiCl ₄ for 10 s
The result shows that the side wall protective radicals in the chamber are reduced and the protective film amount on the pattern side wall is reduced.
Due to the relative increase in the effect of the obliquely incident ions on the side wall protective film, the isolated line pattern has a larger size reduction amount than the internal line pattern, as shown in FIG. The profile of the pattern is more vertical and the dimensional difference between the isolated line pattern and the inner line pattern is reduced.

FIG. 24A shows the results of etching in which Cl ₂ was introduced at 40 sccm, the gas pressure was set to a medium vacuum of 8 Pa, and the exhaust amount was set to 1000 liters / second. As described above, the isolated line pattern has an inversely tapered pattern profile after etching and has a smaller dimension than the internal line pattern. FIG. 24B shows the result of reducing the exhaust amount to 500 liters / sec while keeping the gas pressure constant at 8 Pa. The number of side wall protective radicals in the chamber increases and the amount of protective film on the pattern side wall increases. As the etching effect of the side wall protective film due to the obliquely incident ions is relatively decreased, the dimensional increase amount of the isolated line pattern becomes larger than that of the internal line pattern as shown in FIG. The profile of the inner line pattern becomes more vertical and the dimensional difference between the isolated line pattern and the inner line pattern is reduced.

In FIG. 25 (a), Cl ₂ was introduced at 40 sccm and SiCl ₄ was introduced at 15 sccm, the gas pressure was set to a medium vacuum condition of 10 Pa, and the exhaust amount was set to 1000 l / sec. The results are shown in the figure. As shown in the figure, the isolated line pattern is more
The profile is inversely tapered and the dimensions are narrow. In FIG. 25B, the gas pressure is 10 Pa and the displacement is 1
It shows the result of increasing SiCl ₄ to 25 sccm while keeping it constant at 000 liters / sec. The number of side wall protective radicals in the chamber increases, the amount of protective film on the side wall of the pattern increases, and the side wall protective film by obliquely incident ions is increased. Since the etching effect is relatively small, as shown in the same figure, the isolated line pattern has a larger size increase than the internal line pattern, and the isolated line pattern and the internal line pattern have more vertical profiles. The dimensional difference between the isolated line pattern and the internal line pattern is reduced.

FIG. 26 (a) shows the results obtained by introducing Cl _{2 in an amount} of 40 sccm, setting the gas pressure at a high vacuum of 4 Pa, and performing an evacuation at 1000 liters / sec. As described above, the isolated line pattern has a larger amount of deposited sidewall protective film than the internal line pattern, and the pattern profile after etching has a forward taper and is thicker. FIG. 26 (b) shows
With the gas pressure kept constant at 4 Pa, the displacement is 200
The result was increased to 0 liters / second, the side wall protective radicals in the chamber were decreased, the amount of the protective film on the pattern side wall was decreased, and the etching effect of the side wall protective film by obliquely incident ions was relatively increased. As shown in the figure, the size reduction of the isolated line pattern is larger than that of the internal line pattern, the profiles of the isolated line pattern and the internal line pattern are more vertical, and the space between the isolated line pattern and the internal line pattern is increased. The dimensional difference is reduced.

In FIG. 27 (a), Cl ₂ is introduced at 40 sccm and SiCl ₄ is introduced at 20 sccm, the gas pressure is set to a high vacuum condition of ₄ Pa, and the exhaust amount is set to 1
The result of etching was performed at 000 liters / sec. The isolated line pattern showed a larger amount of deposited sidewall protective film than the internal line pattern, and the pattern profile after etching became a forward taper and the dimension was thick. There is. FIG. 27 (b) shows SiCl while keeping the gas pressure at 4 Pa and the displacement at 1000 l / sec.
_The result of reducing ₄ to 10 sccm shows that the number of side wall protective radicals in the chamber is decreased, the amount of protective film on the pattern side wall is reduced, and the effect of obliquely incident ions on the side wall protective film is relatively increased. As shown in the figure, the isolated line pattern has a larger dimension reduction amount than the internal line pattern, and the profiles of the isolated line pattern and the internal line pattern become more vertical,
The dimensional difference between the isolated line pattern and the internal line pattern is reduced.

FIG. 28 (a) shows a result obtained by introducing Cl ₂ by 40 sccm, setting the gas pressure to a high vacuum condition of 4 Pa, and performing an evacuation at 1000 liters / sec. As described above, the isolated line pattern has an inversely tapered pattern profile after etching and has a smaller dimension than the internal line pattern. FIG. 28B shows the result of reducing the exhaust amount to 500 liters / sec while keeping the gas pressure constant at 4 Pa. The number of side wall protective radicals in the chamber increases and the amount of protective film on the pattern side wall increases. As the etching effect of the side wall protective film due to the obliquely incident ions is relatively decreased, the isolated line pattern has a larger size increase than the internal line pattern as shown in FIG. The profile of the inner line pattern becomes more vertical and the dimensional difference between the isolated line pattern and the inner line pattern is reduced.

In FIG. 29 (a), Cl ₂ is introduced at 40 sccm and SiCl ₄ is introduced at 15 sccm, the gas pressure is set to a high vacuum condition of ₄ Pa, and the exhaust amount is set to 1000 l / sec. The results of the etching are shown. As shown in the figure, the isolated line pattern is more
The profile is inversely tapered and the dimensions are narrow. FIG. 29B shows a gas pressure of 4 Pa and a displacement of 10
00 while maintaining constant in liters / sec., The SiCl ₄ 2
The results are shown in FIG. 5 where the number of side wall protective radicals in the chamber is increased, the amount of protective film on the pattern side wall is increased, and the etching effect of the side wall protective film by obliquely incident ions is relatively reduced. As shown
The dimensional increase of the isolated line pattern is larger than that of the internal line pattern, the profiles of the isolated line pattern and the internal line pattern are more vertical, and the dimensional difference between the isolated line pattern and the internal line pattern is reduced. There is.

In FIG. 30 (a), Cl ₂ is introduced at 40 sccm, the gas pressure is set to a medium vacuum condition of 10 Pa, the exhaust amount is 1000 liter / sec, and the bias power is 30.
The results of etching with 0 watt are shown. As shown in the figure, the isolated line pattern has a larger deposition amount of the sidewall protective film than the internal line pattern, and the pattern profile after etching has a forward taper. And the dimensions are fat. FIG. 30B shows a displacement of 1000
Keeping the gas pressure at 15 Pa while keeping constant at liter / sec.
The result shows that the bias power is increased to 400 watts and the bias power is increased to 400 watts. As shown in the figure, the spread of the angular distribution of the ions increases, and the decrease of the ion energy with the increase of the gas pressure is This is compensated for by the increase in the number of lines, and the etching effect of the side wall protective film by obliquely incident ions becomes relatively large, so that the isolated line pattern has a larger size reduction amount than the internal line pattern. Profile is more vertical, reducing the dimensional difference between isolated line patterns and internal line patterns.

In FIG. 31 (a), Cl ₂ is introduced at 40 sccm, the gas pressure is set to a medium vacuum condition of 10 Pa, the displacement is 1000 liters / sec, and the bias power is 30.
The results of etching with 0 watt are shown. As shown in the figure, the isolated line pattern has a larger deposition amount of the sidewall protective film than the internal line pattern, and the pattern profile after etching has a forward taper. And the dimensions are fat. FIG. 31 (b) shows a gas pressure of 15 P with the bias power kept constant at 300 watts.
The results of increasing the exhaust volume to 2000 liters / sec with increasing to a are shown. The spread of the angular distribution of ions becomes large, the amount of deposition of the side wall protective film decreases, and the effect of etching the side wall protective film by obliquely incident ions is increased. As shown in the same figure, the size of the isolated line pattern is larger than that of the internal line pattern, and the isolated line pattern and the profile of the internal line pattern are more vertical as shown in FIG. The dimensional difference between the pattern and the inner line pattern is reduced.

In FIG. 32 (a), Cl ₂ is introduced at 40 sccm, the gas pressure is set to a high vacuum condition of 4 Pa, the exhaust amount is 1000 liter / sec, and the frequency of the high frequency power is 1
The result of etching at 3.56 MHz is shown. As shown in the figure, the isolated line pattern has an inversely tapered pattern profile after etching and a smaller dimension than the internal line pattern. FIG. 32 (b) shows the result of increasing the frequency of the high-frequency power to 50 MHz while keeping the gas pressure at 4 Pa and the displacement at 1000 liters / sec. As the ratio becomes smaller, the protective film etching amount especially on the side wall of the isolated line pattern decreases. As a result, as shown in the figure, the isolated line pattern has a larger dimension increase amount than the internal line pattern, and the isolated line pattern Also, the profile of the internal line pattern becomes more vertical, and the dimensional difference between the isolated line pattern and the internal line pattern is reduced.

In FIG. 33 (a), Cl ₂ was introduced at 40 sccm, the gas pressure was set to a high vacuum condition of 4 Pa, the exhaust amount was 800 liters / sec, and the frequency of the high frequency power was 50 M.
The results of etching performed at a frequency of Hz are shown. As shown in the figure, the isolated line pattern has a reverse taper pattern profile after etching and a smaller dimension than the internal line pattern. FIG. 33 (b) shows
It shows the result of increasing the frequency of high frequency power to 100MHz while keeping the displacement constant at 800 liters / second, and the gas pressure at which plasma is generated can be stably lowered to 0.5 Pa. As the ratio of incident ions is reduced, the etching amount of the protective film on the side wall of the isolated line pattern is particularly reduced.
As shown in (b), the dimensional increase amount of the isolated line pattern is larger than that of the internal line pattern, the profiles of the isolated line pattern and the internal line pattern are more vertical, and the isolated line pattern and the internal line pattern are The dimensional difference between them is decreasing. Also, set the frequency to 50
A sufficient effect was obtained even if the gas pressure was lowered to 2 Pa while the frequency was kept constant at MHz.

FIG. 34 (a) shows the results obtained by introducing Cl _{2 in an amount} of 40 sccm, setting the gas pressure to a high vacuum of 4 Pa, evacuation rate of 1000 liters / sec, and sample stage temperature of 30 degrees. , And as shown in the figure,
The isolated line pattern is better than the internal line pattern
The deposition amount of the side wall protective film is large, the pattern profile after etching is forward tapered, and the dimension is thick. In FIG. 34 (b), the gas pressure is 4 Pa and the displacement is 10
While keeping constant at 00 liter / sec, set the sample stage temperature to 8
The result shows that the side wall protective radicals deposited on the side wall of the isolated line pattern are reduced, the protective film amount on the pattern side wall is reduced, and the etching effect of the side wall protective film by obliquely incident ions is relatively increased. As a result, as shown in the figure, the dimension reduction amount of the isolated line pattern becomes larger than that of the internal line pattern, the profiles of the isolated line pattern and the internal line pattern become more vertical, and the isolated line pattern and the internal line pattern The dimensional difference between and is decreasing.

FIG. 35 (a) shows a result obtained by introducing Cl _{2 in an amount} of 40 sccm, setting a gas pressure at a medium vacuum of 10 Pa, an exhaust rate of 1000 liter / sec, and a sample stage temperature of 30 ° C. As shown in the figure, the isolated line pattern has a smaller sidewall protective film deposition amount than the internal line pattern, and the pattern profile after etching has a reverse taper and a smaller dimension. FIG. 35B shows the result of lowering the sample stage temperature to 0 degrees while keeping the gas pressure constant at 10 Pa and the exhaust amount constant at 1000 l / sec. As the number of radicals increases, the amount of protective film on the side wall of the pattern increases, and the effect of etching the side wall protective film by obliquely incident ions becomes relatively small, as shown in the figure, the isolated line pattern is better than the internal line pattern. Also, the dimensional increase amount is large, the profiles of the isolated line pattern and the internal line pattern are more vertical, and the dimensional difference between the isolated line pattern and the internal line pattern is reduced.

As another embodiment, Cl ₂ is added for 40 seconds.
ccm is introduced, the gas pressure is 10 Pa under a medium vacuum condition,
When etching was performed at an exhaust rate of 2000 liters / sec and the sample stage temperature of 0 ° C., the amount of deposition of the sidewall protective film on the internal line pattern was smaller than that on the isolated line pattern, and the pattern profile after etching was inversely tapered. However, as a result of setting the gas pressure to 3 Pa, the components of obliquely incident ions become smaller, the profiles of the isolated line pattern and the internal line pattern become more vertical, and the size of the isolated line pattern and the internal line pattern The difference could be reduced.

The embodiment shown in FIGS. 22 to 35 is an example in which a measure for improving the process condition is implemented in order to solve the problem when the process is performed under a certain process condition. However, during the etching process, the internal plasma state is not always constant but changes to some extent with time. For this reason, the ion flux density near the sample stage, the energy of the ions, the incident angle distribution of the ions, and the side wall protecting radical flux also change to some extent with time. By changing the external parameter by the external parameter control device so as to compensate for such a temporal change, the profiles of the isolated line pattern and the internal line pattern are made more vertical, and the isolated line pattern and the internal line pattern are It was possible to reduce the dimensional difference between.

Although the embodiment shown in FIGS. 22 to 35 is carried out in the main etching operation mode,
In the over-etching operation mode in which the underlying oxide film begins to appear, the proportion of reaction products decreases, so the proportion of side wall protecting radicals also decreases.

FIG. 36 shows a state in which etching was carried out by switching between the main etching operation mode and the overetching operation mode using a parallel plate type reactive ion dry etching apparatus, and phosphorus was doped in the same manner as described above. The state of etching performed on the polycrystalline silicon film is shown.

FIG. 36 (a) shows that in the main etching operation mode, Cl ₂ was introduced at 40 sccm, the gas pressure was set to a high vacuum condition of 4 Pa, the exhaust amount was 1000 liter / sec, and the frequency of the high frequency power was 13. The results of etching at 56 MHz are shown below.
The profiles of the isolated line pattern and the internal line pattern are substantially vertical, and the dimensional difference between the isolated line pattern and the internal line pattern is small. Fig. 36
(B) shows the result of over-etching under the above conditions. As shown in the figure, the isolated line pattern has a reverse taper pattern profile after etching and a smaller dimension than the internal line pattern. ing. In order to improve the state of FIG. 36 (b), FIG. 36 (c) shows that the frequency of the high frequency power is 50 MHz while keeping the gas pressure at 4 Pa and the displacement at 1000 liters / sec.
The result shows that the sheath width decreases, the proportion of obliquely incident ions decreases, and as shown in the figure, the protective film etching amount on the side wall of the isolated line pattern decreases.
The profiles of the isolated line pattern and the internal line pattern are more vertical, and the dimensional difference between the isolated line pattern and the internal line pattern is reduced. In this case, in order to increase the etching selection ratio with respect to the silicon oxide film, the power of the high frequency power is reduced to 200 W, thereby reducing the inter-sheath voltage and lowering the ion energy.

In the etching process, the operation mode switching as described above is performed as follows. That is, the etching end point detector 20 is attached to the chamber 11 for generating plasma, and the etching end point detector 20 is attached.
Or a controller capable of pre-programming the switching of the operation mode according to the passage of time is installed, and the signal of this controller is used to determine the end of the main etching. Until the end of etching, perform the etching under the main etching conditions described above,
After the main etching is completed, the etching is performed under the above-mentioned over-etching conditions.

In each of the above-mentioned embodiments, the polycrystalline silicon film is etched. However, even if the present invention is used for etching a metal such as an oxide film, a Si compound or aluminum, or a resist in a multi-layer resist, a high effect can be obtained. Is obtained. When the present invention is applied to etching for aluminum, BCl ₃ + Cl ₂ or S
A gas based on chlorine such as iCl ₄ + Cl ₂ + CHCl ₃ is used, and the gas pressure is 0.1 Pa to 20 Pa.
It is preferable that According to the experiment, the etching rate in this case was 400 to 900 nm / min.

Further, in the above embodiment, the parallel plate reactive ion etching apparatus was used as the apparatus for generating plasma, but instead of this, an electron cyclotron plasma generating apparatus, an electromagnetic induction type plasma generating apparatus or the like is used. In addition, a satisfactory effect can be obtained even in a plasma generator in which the plasma generation power and the bias power can be controlled independently.

An electromagnetic induction type plasma etching apparatus according to another embodiment of the present invention and a dry etching method using the apparatus will be described below with reference to the drawings.

FIG. 37 shows a schematic structure of an electromagnetic induction type plasma etching apparatus used in the plasma generating method according to the present invention. As shown in FIG. 37, a reactive gas is introduced into the metal chamber 11 through the gas controller 12, and the inside of the chamber 11 is controlled to an appropriate pressure by the exhaust system 13.

A spiral coil 25 is provided in the upper part of the chamber 11, and a sample table 15 serving as a cathode is provided in the lower part. The sample stage 15 includes an impedance matching circuit 16 for forming a self DC bias.
The first high frequency power source 17 is connected via. In order to generate plasma in the spiral coil 25,
A second high frequency power source 28 is connected via an impedance matching circuit 27. The frequency of the first high-frequency power source 17 can be changed by the frequency control circuit 21, and the frequency of the second high-frequency power source 28 can be changed by the frequency control circuit 29.

The energy distribution of ions and the width of the sheath region near the sample stage 15 can be determined by the plasma parameter detector 26, and the etching end point can be determined by the etching end point detector 20 using the spectral method. Further, the gas controller 12 and the exhaust system 13 are controlled by a signal from the etching end point detector 20, and the gas pressure and the exhaust amount in the chamber 11 are appropriately controlled. Further, the frequency of the first high frequency power source 17 can be controlled via the frequency control circuit 21 by the signal from the etching end point detector 20.

The temperature of the sample table 15 can be adjusted by controlling the heater 23 via the temperature control circuit 24. In addition, the external parameter control device 22 controls the signal from the plasma parameter detector 26 and the signal from the etching end point detector 20, or the value of these signals and the external parameters such as frequency, gas pressure, high frequency power power, and sample stage temperature. It is possible to control the gas controller 12, the exhaust system 13, the frequency control circuit 21, and the temperature control circuit 24 based on a combination of the above, or a combination of these signals and a pre-programmed processing flow.

A concrete example of the dry etching method performed by using the above-mentioned electromagnetic induction type plasma etching apparatus will be described below with reference to FIG. That is, in FIG. 38 (a), Cl ₂ is introduced by 40 sccm, and 30 is applied to the spiral coil 26 for plasma generation.
0 watt was applied, bias power was 100 watts, gas pressure was high vacuum condition of 3 Pa, and displacement was 1
The results of etching at 000 liters / second are shown. As shown in the figure, the internal line pattern has a larger sidewall protection film deposition amount than the isolated line pattern, and the pattern profile after etching is in order. Tapered and fat. In FIG. 38 (b), the bias power is 1 while the gas pressure is kept constant at 3 Pa.
The result of increasing the displacement to 50 watts and 2000 liters / sec shows that the incident distribution of ions becomes more vertical and the number of side wall protective radicals reaching the side wall of the pattern decreases. As a result, as shown in FIG. The amount of the protective film on the side wall is reduced, the dimension reduction amount of the internal line pattern is larger than that of the isolated line pattern, the profiles of the isolated line pattern and the internal line pattern are more vertical, and the isolated line pattern and the internal line pattern are The dimensional difference between them is decreasing.

Furthermore, even if the gas pressure is reduced to 0.5 Pa, the high frequency power of the bias power is increased from 13.56 MHz to 50 MHz, or the gas pressure is 3 Pa.
From 1 to 1 Pa and the displacement is 200 liters /
The same effect was observed even if it was increased to seconds.

[0215]

According to the plasma generation method of the first aspect of the present invention, the etching amount for the first and second line patterns is increased and the etching amount for the first line pattern is increased for the second line pattern. Since the line width of the first line pattern approaches the line width of the second line pattern, the line widths of the first and second line patterns are controlled so that the line width of the first line pattern becomes smaller than that of the resist pattern. Since the width approaches the width, the line widths of the first and second line patterns are optimized.

According to the plasma generating method of the second aspect of the present invention, since the amount of gas discharged from the vacuum chamber is increased, the amount of the side wall protective film deposited on the side wall of the line pattern is reduced. As the size reduction amount of the second line pattern increases, the etching effect of the obliquely incident ions on the side wall protection film relatively increases.
It is larger than the size reduction of the line pattern.

According to the plasma generating method of the third aspect of the present invention, since the proportion of the sidewall protecting gas in the source gas is reduced, the sidewall protecting radicals in the vacuum chamber are reduced and the sidewalls deposited on the sidewalls of the line pattern. Since the amount of the protective film is reduced, the size reduction amount of the first and second line patterns is increased, and the etching effect of the obliquely incident ions on the side wall protective film is relatively increased, so that the size reduction of the second line pattern is reduced. The amount is larger than the size reduction amount of the first line pattern.

According to the plasma generating method of the fourth aspect of the present invention, since the gas pressure of the raw material gas introduced into the vacuum chamber is increased and the power of the high frequency power is increased, the ion energy is reduced as the gas pressure is increased. Is compensated by the increase of the high frequency power and the etching effect on the side wall protective film by the obliquely incident ions is increased, so that the dimensional reduction amount in the first and second line patterns is increased and the spread of the angle distribution of the incident ions is increased. As compared with the second line pattern, the etching effect of the first line pattern on the sidewall protection film is relatively reduced, so that the size reduction amount of the first line pattern is smaller than the size reduction amount of the second line pattern. To do.

According to the plasma generating method of the fifth aspect of the present invention, since the gas pressure of the raw material gas introduced into the vacuum chamber is increased and the discharge amount of the gas exhausted from the vacuum chamber is increased, the side wall inside the vacuum chamber is increased. Since the protective radicals are reduced and the amount of the sidewall protective film deposited on the sidewalls of the line pattern is reduced, the dimensional reduction amount in the first and second line patterns is increased and the spread of the angle distribution of incident ions is increased. Since the etching effect of the first line pattern on the sidewall protection film is relatively reduced as compared with the second line pattern, the dimension reduction amount of the first line pattern is smaller than the dimension reduction amount of the second line pattern. .

According to the plasma generating method of the sixth aspect of the present invention, since the temperature of the sample stage is raised, the side wall protecting radicals deposited on the side wall of the line pattern are reduced, and the size reduction in the first and second line patterns is reduced. As the amount increases, the etching effect of the obliquely incident ions on the side wall protection film increases, so that the dimension reduction amount of the second line pattern increases more than the dimension reduction amount of the first line pattern.

According to the plasma generating method of the seventh aspect of the present invention, the first and second parameters are controlled in the parameter control step.
The line width of the first line pattern is controlled to be smaller than the line width of the first line pattern because the etching amount for the first line pattern is controlled to be relatively smaller than the etching amount for the first line pattern. Since the line widths of the first and second line patterns approach the line width of the resist pattern while approaching the line width of the second line pattern, the line widths of the first and second line patterns are optimized.

According to the plasma generating method of the eighth aspect of the present invention, since the gas pressure of the source gas introduced into the vacuum chamber is reduced, the obliquely incident ions are reduced and the effect of etching the side wall protective film by the obliquely incident ions is reduced. Since it is reduced, the dimensional reduction amount in the first and second line patterns is reduced.

According to the plasma generation method of the ninth aspect of the present invention, since the amount of gas discharged from the vacuum chamber is reduced, the side wall protection radicals in the vacuum chamber increase and the side wall protection deposited on the side wall of the line pattern. Since the amount of the film is increased, the dimensional reduction amount in the first and second line patterns is reduced, and the reduction of the etching effect on the side wall protective film by the obliquely incident ions is remarkable in the second line pattern. The dimension reduction amount of the second line pattern is smaller than the dimension reduction amount of the first line pattern.

According to the plasma generating method of the tenth aspect of the present invention, since the temperature of the sample stage is raised, the side wall protecting radicals deposited on the side walls of the line pattern are reduced, and the size reduction in the first and second line patterns is reduced. As the amount increases, the etching effect of the obliquely incident ions on the side wall protection film increases, so that the dimension reduction amount of the second line pattern increases more than the dimension reduction amount of the first line pattern.

According to the plasma generating method of the eleventh aspect of the present invention, in the parameter control step, the etching amount for the first and second line patterns is increased and the etching amount for the first line pattern is set to the second amount.
Since the etching amount is controlled to be increased relative to the line pattern of the first line pattern, the line width of the first line pattern approaches the line width of the second line pattern, and the lines of the first and second line patterns are Since the width approaches the line width of the resist pattern, the line widths of the first and second line patterns are optimized.

According to the plasma generating method of the invention of claim 12, since the high frequency power is increased and the amount of gas discharged from the vacuum chamber is increased, the sidewall protective radicals in the vacuum chamber are reduced. As a result, the amount of the side wall protective film deposited on the side wall of the line pattern is reduced, so that the dimensional reduction amount in the first and second line patterns is increased and the spread of the angular distribution of the incident ions is reduced, and the first Since the etching effect of the line pattern on the sidewall protection film is relatively increased, the size reduction amount of the first line pattern is larger than the size reduction amount of the second line pattern.

According to the plasma generating method of the thirteenth aspect of the invention, the gas pressure of the source gas introduced into the vacuum chamber is reduced, so that the obliquely incident ions are reduced and the effect of etching the side wall protective film by the obliquely incident ions is reduced. Since the reduction remarkably appears in the second line pattern, the dimensional reduction amount of the second line pattern is relatively reduced.

According to the plasma generating method of the fourteenth aspect of the present invention, the gas pressure of the raw material gas introduced into the vacuum chamber is reduced and the amount of the gas exhausted from the vacuum chamber is increased. Since the protective radicals are reduced and the amount of the sidewall protective film deposited on the sidewalls of the line pattern is reduced, the size reduction amount in the first and second line patterns is increased and the spread of the angular distribution of incident ions is reduced. As a result, the etching effect of the first line pattern on the side wall protection film is relatively increased, so that the size reduction amount of the first line pattern is larger than the size reduction amount of the second line pattern.

According to the plasma generating method of the fifteenth aspect of the present invention, since the frequency of the high frequency power is increased, the sheath width is reduced and the ratio of obliquely incident ions is decreased, so that the etching effect of the obliquely incident ions on the side wall protective film is improved. Since the reduction significantly appears in the second line pattern, the dimension reduction amount of the second line pattern is relatively reduced.

According to the plasma generating method of the sixteenth aspect of the present invention, since the temperature of the sample stage is raised, the side wall protecting radicals deposited on the side walls of the line pattern are reduced, and the size of the first and second line patterns is reduced. As the amount increases, the etching effect of the obliquely incident ions on the side wall protection film increases, so that the dimension reduction amount of the second line pattern increases more than the dimension reduction amount of the first line pattern.

According to the plasma generating method of the seventeenth aspect, in the parameter control step, the etching amount for the first and second line patterns is reduced and the etching amount for the first line pattern is set to the second amount.
Since the etching amount is controlled to be increased relative to the line pattern of the first line pattern, the line width of the first line pattern approaches the line width of the second line pattern, and the lines of the first and second line patterns are Since the width approaches the line width of the resist pattern, the line widths of the first and second line patterns are optimized.

According to the plasma generating method of the eighteenth aspect of the present invention, in order to reduce the discharge amount of the gas discharged from the vacuum chamber, the side wall protective radicals in the vacuum chamber increase and the side wall protective deposits on the side wall of the line pattern. Since the amount of the film is increased, the dimensional reduction amount in the first and second line patterns is reduced, and the reduction of the etching effect on the side wall protective film by the obliquely incident ions is remarkable in the second line pattern. The dimension reduction amount of the second line pattern is smaller than the dimension reduction amount of the first line pattern.

According to the plasma generating method of the nineteenth aspect of the present invention, since the ratio of the side wall protecting gas in the source gas is increased, the side wall protecting radicals in the vacuum chamber are increased and the side walls are deposited on the side walls of the line pattern. Since the amount of the protective film increases, the dimensional reduction amount in the first and second line patterns is reduced, and the reduction of the etching effect on the sidewall protective film due to the obliquely incident ions remarkably appears in the second line pattern. The dimension reduction amount of the second line pattern is smaller than the dimension reduction amount of the first line pattern.

According to the plasma generating method of the twentieth aspect of the present invention, since the frequency of the high frequency power is increased, the sheath width is reduced and the ratio of obliquely incident ions is decreased. Is significantly exhibited in the second line pattern, so that the dimensional reduction amount of the second line pattern is relatively reduced.

According to the plasma generating method of the twenty-first aspect of the present invention, since the frequency of the high frequency power is increased and the gas pressure of the raw material gas introduced into the vacuum chamber is reduced, the obliquely incident ions are greatly reduced and the oblique incidence is performed. Since the etching effect on the sidewall protection film by the ions is reduced, the size reduction amount in the first and second line patterns is reduced, and the etching effect on the sidewall protection film by the obliquely incident ions is significantly reduced in the second line pattern. , The dimension reduction amount of the second line pattern is smaller than the dimension reduction amount of the first line pattern.

According to the plasma generating method of the twenty-second aspect of the invention, since the temperature of the sample stage is lowered, the amount of the side wall protective film deposited on the side wall of the line pattern increases, so that the first and second line patterns are formed. In addition to reducing the size reduction amount of the second line pattern, the size reduction amount of the second line pattern is smaller than that of the first line pattern.

According to the plasma generating method of the twenty-third aspect of the present invention, since the gas pressure of the source gas introduced into the vacuum chamber is reduced, the obliquely incident ions are reduced, and the effect of etching the side wall protective film by the obliquely incident ions is reduced. Since the amount of reduction in the size of the first and second line patterns is reduced, the reduction of the etching effect on the side wall protective film by the obliquely incident ions appears remarkably in the second line pattern. Of the first line pattern is smaller than that of the first line pattern.

[Brief description of drawings]

FIG. 1 is a schematic view of a dry etching apparatus used in a dry etching method according to each embodiment of the present invention.

FIG. 2 is a schematic diagram showing a state in which ions starting from a boundary between a bulk plasma region and a sheath region are transported toward a sample stage while colliding with neutral particles in the sheath region.

FIG. 3 shows the characteristics of the incident distribution of ions starting from the boundary between the bulk plasma region and the sheath region when the gas pressure is 0.1 Pa, (a) shows the ion angle distribution,
(B) shows the ion energy distribution.

FIG. 4 shows the characteristics of the incident distribution of ions starting from the boundary between the bulk plasma region and the sheath region when the gas pressure is 0.2 Pa, (a) shows the ion angle distribution,
(B) shows the ion energy distribution.

FIG. 5 shows the characteristics of the incident distribution of ions starting from the boundary between the bulk plasma region and the sheath region when the gas pressure is 0.5 Pa, (a) shows the ion angle distribution,
(B) shows the ion energy distribution.

FIG. 6 shows the characteristics of the incident distribution of ions starting from the boundary between the bulk plasma region and the sheath region when the gas pressure is 1.0 Pa, (a) shows the ion angle distribution,
(B) shows the ion energy distribution.

FIG. 7 shows the characteristics of the incident distribution of ions starting from the boundary between the bulk plasma region and the sheath region when the gas pressure is 2.0 Pa, (a) shows the ion angle distribution,
(B) shows the ion energy distribution.

FIG. 8 shows the characteristics of the incident distribution of ions starting from the boundary between the bulk plasma region and the sheath region when the gas pressure is 3.0 Pa, (a) shows the ion angle distribution,
(B) shows the ion energy distribution.

FIG. 9 shows the characteristics of the incident distribution of ions starting from the boundary between the bulk plasma region and the sheath region when the gas pressure is 5.0 Pa, (a) shows the ion angle distribution,
(B) shows the ion energy distribution.

FIG. 10 shows the characteristics of the incident distribution of ions starting from the boundary between the bulk plasma region and the sheath region when the gas pressure is 10 Pa, (a) shows the ion angle distribution,
(B) shows the ion energy distribution.

FIG. 11 is a diagram showing the gas pressure dependence of the standard deviation σ representing the spread of the scattering angle in the incident angle distribution of ions starting from the boundary between the bulk plasma region and the sheath region and the mean free path λ of the ions.

FIG. 12 shows a finite scattering angle ± in the angular distribution of ions starting from the boundary between the bulk plasma region and the sheath region.
It is a figure which shows the gas pressure dependence of the ratio with respect to the number of all the incident ions which inject within Δ (degree).

FIG. 13 is a diagram showing the behavior of sidewall protecting radicals that are isotropically incident from above the line-and-space pattern.

FIG. 14 is a diagram showing a mechanism of a difference in profile and dimensional change in an internal line pattern and an isolated line pattern of a line and space pattern.

FIG. 15 is a diagram illustrating how internal parameters change in a vacuum chamber when external operating parameters including gas pressure and exhaust amount are changed.

FIG. 16 is a diagram for explaining how the internal parameters in the vacuum chamber change when the external operating parameters including bias power are changed.

FIG. 17 is a diagram illustrating a method of reducing the dimensional difference between an internal line pattern and an isolated line pattern in the case of medium vacuum.

FIG. 18 is a diagram illustrating a method of reducing a dimensional difference between an internal line pattern and an isolated line pattern in the case of high vacuum.

FIG. 19 is a diagram for explaining first means to reduce the dimensional difference between the internal line pattern and the isolated line pattern in the case of medium vacuum and high vacuum.

FIG. 20 illustrates how to deal with the cases (E) to (H) shown in FIG. 19 after that in order to reduce the dimensional difference between the internal line pattern and the isolated line pattern. It is a figure.

FIG. 21 is a cross-sectional view showing a line pattern before dry etching.

FIG. 22 (a) shows that Cl ₂ is introduced at 40 sccm, the gas pressure is 10 Pa, and the exhaust pressure is 1000 Pa.
When etching is performed at a rate of 1 liter / sec, the pattern profile after etching shows a forward taper and the dimension is thick, and (b) shows a gas pressure of 10P.
FIG. 6 is a diagram showing a state in which the pattern profile is improved by increasing the exhaust amount to 2000 liters / sec while keeping constant at a.

FIG. 23 (a) shows that etching was performed under the conditions of Cl ₂ of 40 sccm and sidewall protecting radicals of SiCl ₄ of 20 sccm, a gas pressure of 10 Pa and a vacuum of 1500 liter / sec. At this time, the pattern profile after etching shows a forward taper and the dimension is thick, and (b) shows a gas pressure of 10P.
a, SiCl ₄ was reduced to 10 sccm while keeping the displacement at 1500 liters / sec.
It is a figure which shows the state which improved the profile.

FIG. 24 (a) shows a pattern profile after etching when etching was performed with Cl ₂ of 40 sccm introduced, a gas pressure of 8 Pa and a medium vacuum, and a displacement of 1000 l / sec. FIG. 6B is a diagram showing a state in which the taper is inversely tapered and the dimensions are thin, and FIG. 7B is a diagram showing a state where the pattern profile is improved by reducing the exhaust amount to 500 liters / sec while keeping the gas pressure constant at 8 Pa. .

FIG. 25 (a) shows Cl ₂ at 40 sccm and SiCl ₄
Of 15 sccm and a gas pressure of 10 Pa and a vacuum of 10 liters / second, the pattern profile after etching becomes a reverse taper and the dimension is thin. (B) shows 25 sc of SiCl ₄ with the gas pressure kept at 10 Pa and the displacement at 1000 l / sec.
It is a figure which shows the state which increased to cm and improved the pattern profile.

FIG. 26 (a) shows a pattern profile after etching when etching was performed under the conditions of a high vacuum of Cl ₂ at a gas pressure of 4 Pa and a displacement of 1000 l / sec. FIG. 3B is a diagram showing a state in which the pattern is forward tapered and the size is thick, and FIG. 6B is a diagram showing a state where the pattern profile is improved by increasing the exhaust amount to 2000 liters / sec while keeping the gas pressure constant at 4 Pa.

FIG. 27 (a) shows Cl ₂ at 40 sccm and SiCl ₄
20 sccm was introduced, and when the etching was performed at a gas pressure of 4 Pa and a high vacuum of 4 Pa and an evacuation rate of 1000 liters / second, the pattern profile after etching became a forward taper and the dimension was thick. ,
(B) is a diagram showing a state in which the pattern profile is improved by reducing SiCl ₄ to 10 sccm while keeping the gas pressure at 4 Pa and the exhaust amount at 1000 liter / sec.

FIG. 28 (a) shows a pattern profile after etching when etching was performed under the conditions of a high vacuum of 4 Pa of Cl ₂ and a gas pressure of 4 Pa and a displacement of 1000 liter / sec. FIG. 6B is a diagram showing a state in which the taper is inversely tapered and the dimensions are narrowed, and FIG. 6B is a diagram showing a state in which the gas displacement is reduced to 500 liters / sec and the pattern profile is improved while the gas pressure is kept constant at 4 Pa. .

FIG. 29 (a) shows Cl ₂ at 40 sccm and SiCl ₄
Of 15 sccm and a gas pressure of 4 Pa under a high vacuum condition and an evacuation rate of 1000 liters / second. When etching is performed, the pattern profile after etching becomes a reverse taper and the dimension becomes small. Shows,
(B) is a diagram showing a state where the pattern profile is improved by increasing SiCl ₄ to 25 sccm while keeping the gas pressure at 4 Pa and the exhaust amount at 1000 l / sec.

FIG. 30 (a) shows that Cl ₂ of 40 sccm is introduced, the gas pressure is 10 Pa, and the exhaust pressure is 1000 Pa.
When etching was performed with a bias power of 300 watts per liter / second, the pattern after etching
The profile shows a forward taper and a large size. (B) shows that the gas pressure is increased to 15 Pa and the bias power is increased to 400 watts while keeping the displacement constant at 1000 liter / sec. It is a figure which shows the state which improved the pattern profile.

FIG. 31 (a) shows that Cl ₂ of 40 sccm is introduced, the gas pressure is 10 Pa, and the exhaust pressure is 1000 Pa.
When etching was performed with a bias power of 300 watts per liter / second, the pattern after etching
The profile shows a forward taper and a large size. (B) shows that the gas pressure is increased to 15 Pa and the displacement is increased to 2000 liters / sec while the bias power is kept constant at 300 watts. It is a figure which shows the state which improved the pattern profile.

FIG. 32 (a) shows the case where 40 sccm of Cl ₂ is introduced, the gas pressure is 4 Pa, and the exhaust pressure is 1000 liter / sec, and the frequency of the high frequency power is 13.56 MHz when the etching is performed. The pattern profile after etching shows an inverse taper and the dimensions are thin. (B) shows a gas pressure of 4 Pa and an exhaust amount of 10
It is a figure which shows the state which improved the pattern profile by increasing the frequency of high frequency electric power to 50 MHz, keeping it constant at 00 liters / second.

FIG. 33 (a) shows that when etching is performed with Cl ₂ of 40 sccm introduced, a gas pressure of 4 Pa and a high vacuum, a displacement of 800 liters / sec and a high-frequency power frequency of 50 MHz. The pattern profile after etching shows an inverse taper and the dimension is small. (B) shows the pattern profile by increasing the frequency of the high frequency power to 100 MHz while keeping the displacement at 800 liters / sec. It is a figure which shows the state which improved.

FIG. 34 (a) shows that when etching is performed with Cl ₂ of 40 sccm introduced, a gas pressure of 4 Pa and a high vacuum, a pumping rate of 1000 liters / sec and a sample stage temperature of 30 ° C. The pattern profile after etching shows a forward taper and the dimension is thick, (b)
FIG. 4 is a diagram showing a state in which the pattern profile is improved by increasing the sample stage temperature to 80 degrees while keeping the gas pressure constant at 4 Pa and the exhaust volume constant at 1000 l / sec.

FIG. 35 (a) shows that Cl ₂ of 40 sccm is introduced, the gas pressure is 10 Pa, and the exhaust pressure is 1000 Pa.
When etching was carried out at a sample stage temperature of 30 ° C. in liters / second, the pattern profile after etching shows a reverse taper and the dimension is narrow,
FIG. 6B is a diagram showing a state in which the pattern profile is improved by decreasing the sample stage temperature to 0 degrees while keeping the gas pressure at 10 Pa and the displacement at 1000 liters / sec.

FIG. 36 shows a state in which etching is performed by switching between a main etching operation mode and an overetching operation mode in forming a phosphorus-doped polycrystalline silicon gate by using a parallel plate type reactive ion dry etching apparatus. ) Shows the etching state under the main etching conditions, (b) shows the etching state under the over-etching conditions, and (c) shows the etching state under the improved over-etching conditions.

FIG. 37 is a schematic view of another dry etching apparatus used in the plasma generating method according to the present invention.

38 is a state when a phosphorus-doped polycrystalline silicon gate is formed by using the dry etching apparatus shown in FIG. 37, (a) introducing Cl ₂ of 40 sccm, with respect to a spiral coil for plasma generation. 300 watts, bias power 1000 watts, gas pressure 3 Pa
When etching was performed under the conditions of high vacuum of 100 liter / sec.
The profile shows a forward taper and the dimension is thick. (B) shows a bias power of 150 watts and a displacement of 2000 with the gas pressure kept constant at 3 Pa.
It is a figure which shows the state which increased to 1 liter / second and improved the pattern profile.

[Explanation of symbols]

 11 Metallic Chamber 12 Gas Controller 13 Exhaust System 14 Anode (Anode) 15 Sample Stage (Cathode) 16 Impedance Matching Circuit 17 High Frequency Power Supply (First High Frequency Power Supply) 20 Etching End Point Detector 21 Frequency Control Circuit 22 External Operating Parameter Control Device 23 heater 24 sample stage temperature control device 25 spiral coil 26 plasma parameter detector 27 impedance matching circuit 28 second high frequency power supply 29 frequency control circuit 30 photoresist pattern 31 phosphorus-doped polycrystalline silicon film 32 thermal oxide film 33 silicon substrate 34 Side wall protective deposited film such as reaction product 35 Ion 36 Side wall protective deposited film such as etched reaction product 40 Side wall protected radical

Claims (23)

[Claims] 1. An etching gas for etching a sample to be etched placed on the sample table and having a resist pattern formed on the surface thereof in a vacuum chamber having a sample table at the bottom and the sample to be etched. Introduce a raw material gas consisting of a sidewall protecting gas that generates sidewall protecting radicals that protect the sidewalls of the line pattern formed by etching, and generate ions of the source gas, and at the same time, to the sample stage. A dry etching method of inducing the ions to the sample stage by applying high frequency power to form a self DC bias, thereby performing etching on the sample to be etched, which are formed close to each other. The line pattern located inside the line pattern group consisting of the plurality of line patterns The line width of the first line pattern formed of a line is the outermost line pattern in the line pattern group or the second line pattern formed of the line pattern isolated from the line pattern group. When the line width of the first and second line patterns is thinner than the line width of the resist pattern, the etching amount on the sidewalls of the first and second line patterns is increased. And the gas pressure of the source gas introduced into the vacuum chamber and the vacuum so that the etching amount on the side wall of the first line pattern is relatively reduced compared to the etching amount on the side wall of the second line pattern. Amount of gas discharged from the chamber, frequency of the high frequency power, high frequency power Dry etching method, which comprises a parameter control step of changing at least one parameter of a parameter group consisting of the electric power, the ratio of the sidewall protecting gas in the raw material gas, and the temperature of the sample stage. . 2. The dry etching method according to claim 1, wherein the parameter control step includes a step of increasing an exhaust amount of gas exhausted from the vacuum chamber. 3. The dry etching method according to claim 1, wherein the parameter controlling step includes a step of reducing a ratio of the sidewall protecting gas in the raw material gas. 4. The parameter controlling step includes a step of increasing a gas pressure of a raw material gas introduced into the vacuum chamber and a step of increasing power of the high frequency power. Item 3. The dry etching method according to item 1. 5. The parameter control step includes a step of increasing a gas pressure of a raw material gas introduced into the vacuum chamber, and a step of increasing a discharge amount of a gas discharged from the vacuum chamber. The dry etching method according to claim 1, wherein: 6. The dry etching method according to claim 1, wherein the parameter control step includes a step of increasing the temperature of the sample stage. 7. An etching gas for etching a sample to be etched placed on the sample table and having a resist pattern formed on the surface thereof in a vacuum chamber having a sample table at the bottom and the sample to be etched. Introduce a raw material gas consisting of a sidewall protecting gas that generates sidewall protecting radicals that protect the sidewalls of the line pattern formed by etching, and generate ions of the source gas, and at the same time, to the sample stage. A dry etching method of inducing the ions to the sample stage by applying high frequency power to form a self DC bias, thereby performing etching on the sample to be etched, which are formed close to each other. The line pattern located inside the line pattern group consisting of the plurality of line patterns The line width of the first line pattern formed of a line is the outermost line pattern in the line pattern group or the second line pattern formed of the line pattern isolated from the line pattern group. When the line widths of the first and second line patterns are narrower than the line widths of the resist patterns, the etching amount of the sidewalls of the first and second line patterns is reduced. And the gas pressure of the source gas introduced into the vacuum chamber and the vacuum so that the etching amount on the side wall of the first line pattern is relatively reduced compared to the etching amount on the side wall of the second line pattern. Amount of gas discharged from the chamber, frequency of the high frequency power, high frequency power Dry etching method, which comprises a parameter control step of changing at least one parameter of a parameter group consisting of the electric power, the ratio of the sidewall protecting gas in the raw material gas, and the temperature of the sample stage. . 8. The dry etching method according to claim 7, wherein the parameter control step includes a step of reducing a gas pressure of a raw material gas introduced into the vacuum chamber. 9. The dry etching method according to claim 7, wherein the parameter control step includes a step of reducing the amount of gas discharged from the vacuum chamber. 10. The dry etching method according to claim 7, wherein the parameter control step includes a step of increasing the temperature of the sample stage. 11. An etching gas for etching a sample to be etched placed on the sample table and having a resist pattern formed on the surface thereof in a vacuum chamber having a sample table at the bottom and the sample to be etched are Introduce a raw material gas consisting of a sidewall protecting gas that generates sidewall protecting radicals that protect the sidewalls of the line pattern formed by etching, and generate ions of the source gas, and at the same time, to the sample stage. A dry etching method of inducing the ions to the sample stage by applying high frequency power to form a self DC bias, thereby performing etching on the sample to be etched, which are formed close to each other. The line located inside the line pattern group consisting of the plurality of line patterns The line width of the first line pattern formed of turns is the outermost line pattern in the line pattern group or the line of the second line pattern formed of the line pattern isolated from the line pattern group. When the line width of the first and second line patterns becomes thicker than the width and the line width of the first and second line patterns becomes thicker than the line width of the resist pattern, the etching amount on the sidewalls of the first and second line patterns increases. At the same time, the gas pressure of the source gas introduced into the vacuum chamber and the vacuum chamber so that the etching amount on the side wall of the first line pattern is relatively increased compared to the etching amount on the side wall of the second line pattern. Amount of gas discharged from the, the frequency of the high frequency power, the high frequency power Dry etching, comprising a parameter control step of changing at least one parameter of a parameter group consisting of power of power, ratio of the sidewall protecting gas in the raw material gas, and temperature of the sample stage Method. 12. The parameter controlling step includes a step of increasing the electric power of the high frequency power and a step of increasing an exhaust amount of gas exhausted from the vacuum chamber. The dry etching method described in. 13. The dry etching method according to claim 11, wherein the parameter controlling step includes a step of reducing a gas pressure of a raw material gas introduced into the vacuum chamber. 14. The dry etching method according to claim 13, wherein the parameter controlling step further includes a step of increasing an exhaust amount of gas exhausted from the vacuum chamber. 15. The dry etching method according to claim 11, wherein the parameter controlling step includes a step of increasing a frequency of the high frequency power. 16. The dry etching method according to claim 11, wherein the parameter control step includes a step of increasing the temperature of the sample stage. 17. An etching gas for etching a sample to be etched placed on the sample table and having a resist pattern formed on the surface thereof in a vacuum chamber having a sample table at the bottom and the sample to be etched. Introduce a raw material gas consisting of a sidewall protecting gas that generates sidewall protecting radicals that protect the sidewalls of the line pattern formed by etching, and generate ions of the source gas, and at the same time, to the sample stage. A dry etching method of inducing the ions to the sample stage by applying high frequency power to form a self DC bias, thereby performing etching on the sample to be etched, which are formed close to each other. The line located inside the line pattern group consisting of the plurality of line patterns The line width of the first line pattern formed of turns is the outermost line pattern in the line pattern group or the line of the second line pattern formed of the line pattern isolated from the line pattern group. When the line width of the first and second line patterns becomes narrower than the width and the line width of the first and second line patterns becomes narrower than the line width of the resist pattern, the etching amount on the sidewalls of the first and second line patterns decreases. At the same time, the gas pressure of the source gas introduced into the vacuum chamber and the vacuum chamber so that the etching amount on the side wall of the first line pattern is relatively increased compared to the etching amount on the side wall of the second line pattern. Amount of gas discharged from the, the frequency of the high frequency power, the high frequency power Dry etching, comprising a parameter control step of changing at least one parameter of a parameter group consisting of power of power, ratio of the sidewall protecting gas in the raw material gas, and temperature of the sample stage Method. 18. The dry etching method according to claim 17, wherein the parameter control step includes a step of reducing the amount of gas discharged from the vacuum chamber. 19. The dry etching method according to claim 17, wherein the parameter controlling step includes a step of increasing a ratio of the sidewall protecting gas in the raw material gas. 20. The dry etching method according to claim 17, wherein the parameter controlling step includes a step of increasing a frequency of the high frequency power. 21. The parameter controlling step further comprises a step of reducing a gas pressure of a raw material gas introduced into the vacuum chamber.
The dry etching method described in. 22. The dry etching method according to claim 17, wherein the parameter control step includes a step of lowering the temperature of the sample stage. 23. The dry etching method according to claim 17, wherein the parameter control step includes a step of reducing a gas pressure of a raw material gas introduced into the vacuum chamber.