JPH09236904A - Photomask - Google Patents

Abstract

PROBLEM TO BE SOLVED: To increase the depth of a focus and to prevent the reduction of the thickness of a film and the vanishment of the film. SOLUTION: Patterns having periodicity are formed on a transparent substrate 1. The outermost ones of the patterns are halftone patterns 13 and the remainders are light shielding patterns 12. The halftone patterns 13 are formed with a translucent film 3 and the light shielding patterns 12 are formed with a laminated film consisting of a translucent film 3 and a light shielding film 2. The extent of phase shift of the translucent film 3 is regulated to 180 deg..

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a photomask used in a lithography process for an exposure apparatus, and more particularly to a photomask having a periodic pattern.

[0002]

2. Description of the Related Art At present, in a manufacturing process of a semiconductor device, an optical lithography technique is mainly used for forming a pattern on a semiconductor substrate. In optical lithography,
A pattern of a photomask (which is an exposure master plate on which a pattern composed of a transparent region and a light-shielded region is formed, and is also called a reticle when the reduction ratio is not 1: 1 is referred to as a photomask in this case) by an exposure device. It is transferred onto a semiconductor substrate coated with a photosensitive resin and developed to form a predetermined pattern on the photosensitive resin. The pattern of this photosensitive resin serves as a mask in etching or impurity introduction (ion implantation) which is the next step of lithography (this photosensitive resin is also called a resist in the sense of resisting etching).

In the conventional photolithography technology, development of an exposure apparatus has been mainly carried out, and in particular, miniaturization of a semiconductor element pattern has been dealt with by increasing the NA of a projection lens system. Here, NA (numerical aperture) corresponds to how much light the lens can collect, and the larger this value, the more light can be collected, and the better the lens performance. Further, as is generally well known as the Rayleigh equation, the limit resolution R (dimension of the fine pattern of the limit that can be resolved) and NA are R = K ₁ λ / NA (where K ₁ Is a constant that depends on the process such as the performance of the photosensitive resin), and the larger the NA, the higher the limit resolution and the better miniaturization.

However, although the resolution is improved by increasing the NA of the exposure apparatus, on the contrary, the depth of focus (the range in which the shift of the focus position can be tolerated) decreases, and further miniaturization becomes difficult in terms of the depth of focus. ing. Here again, the actual physical explanation is omitted, but the depth of focus DOF
It is known that the relation of DOF = K ₂ × λ / NA ² (here, K ₂ is a constant depending on the process) is established between and NA. That is, as the NA is increased, the depth of focus becomes narrower, and a slight shift of the focal position cannot be tolerated.

Accordingly, various super-resolution techniques have been studied. In general, the super-resolution method is a method of improving the light intensity distribution on the image plane by controlling the transmittance and phase of light on the pupil plane of the illumination optical system, photomask, and projection lens system. Further, among various super-resolution methods, improvement of the resolution characteristics by optimizing the illumination optical system, so-called modified illumination method is highly feasible and has been particularly attracting attention in recent years. Before describing the modified illumination method, first, a brief description will be given of an illumination optical system of a normal exposure apparatus (generally called a stepper). In a lithography process for forming a semiconductor element, it is necessary to illuminate the entire exposed area of a photomask with a uniform intensity in order to control the dimension of a formed pattern on the exposed area of a semiconductor substrate. There is. Therefore, light emitted from a mercury lamp, which is a light source, is converted into a single wavelength through a cold mirror, an interference filter, and the like, and then guided to a fly-eye lens that is an optical element for obtaining illuminance uniformity. The fly-eye lens is an optical element in which a plurality of single lenses of the same type are bundled in parallel, and each single lens of the fly-eye lens forms an independent point light source with a focus, and this point light source group illuminates the photomask. This improves the illuminance uniformity on the photomask.

This point light source group is sometimes called a secondary light source, while the mercury lamp is called a primary light source. Also, when illuminating through the fly-eye lens like this,
The light emission state of the mercury lamp, which is the original light source, has nothing to do with the illumination state of the photomask. Only the shape and the intensity distribution of the point light source formed by the fly-eye lens substantially determine the illumination state of the photomask and affect the exposure characteristics. Therefore, this point light source group is also called an effective light source. Generally, this effective light source has a circular shape, and its size is represented by a coherent factor σ (σ = NA of the illumination optical system / NA of the projection lens system). Before,
This σ value was fixed and was about 0.5 to 0.7, but with the miniaturization of the pattern size, the σ value was made variable, and the optimum value was set within the range of 0.3 to 0.8 for each pattern. I have come to choose.

Further, it has been proposed to control the shape of this effective light source to improve the resolution characteristics. This is a method generally called the modified illumination method. As means for changing the shape of the effective light source, usually, a diaphragm or a filter of various shapes is arranged immediately after the fly-eye lens.
Note that this method is distinguished by the shape of the effective light source (shape of the diaphragm). For example, an illumination method that shields the central portion of the diaphragm and uses a ring-type illumination light source is called an annular illumination method.

Next, the effect of this modified illumination method will be briefly described. FIG. 7 shows a comparison of the illumination states of normal illumination and ring illumination. In a normal illumination method, a diaphragm 102a having a circular aperture as shown in FIG. 7A is used.
As shown in (b), there is light that enters the photomask 6 almost vertically. In order to resolve the pattern of the photomask 6, it is necessary to collect at least 0th order light and + 1st or -1st order light out of the diffracted light, but when the pattern becomes fine, the diffraction angle becomes large and the projection It will not fit in the lens 103 system.

Therefore, in the fine pattern, the vertically incident light becomes a noise component that does not contribute to resolution,
This lowers the contrast of the light intensity distribution on the image plane. However, when the diaphragm 102b having a ring-shaped opening as shown in FIG. 7C is used, light is incident on the photomask 6 only obliquely, and the + 1st or -1st order diffracted light is projected accordingly. Now comes into the lens 103,
Most of the illumination light comes to help resolve the pattern. As described above, the resolution and the depth of focus can be improved by removing the vertically incident component of the illumination light that does not contribute to the resolution.

In addition to the modified illumination method, a phase shift mask, which is a super-resolution method by improving the photomask side, has been actively studied. As a phase shift mask, called Shibuya-Levenson method,
A method was first proposed in which the phase of light passing through a transparent region in a periodic pattern is alternately changed by 180 degrees. However, this method has a problem that it is effective only for periodic patterns and cannot be applied to isolated patterns. Therefore, other methods such as an auxiliary pattern or a halftone method have been proposed thereafter. Among them, the halftone type phase shift mask is particularly attracting attention because it is easier to design and manufacture the mask than other methods.

Before describing these phase shift masks, a normal photomask is first shown in FIG. 8 for comparison. 8A is a plan view and FIG. 8B is a cross-sectional view taken along the line AA. With a normal mask, as shown in FIG.
As shown in (b), 70 to 1 is formed on the transparent substrate 1 such as quartz.
00nm thick chromium (Cr) and chromium oxide (Cr
The light shielding film 2 of (O) is formed. Then, FIG.
As shown in, the amplitude of the light transmitted through the mask becomes a constant value at the opening (transparent part) and becomes zero at the other part (light shielding part). Generally, an image formed through an optical system is described by Fourier transform.However, a mask pattern illuminated by transmitted illumination and subjected to Fourier transform is subjected to Fourier inverse transform by a projection lens system, and the original mask pattern is again formed on the image plane. It is formed. However, at this time, since the projection lens system functions as a low-pass filter, the higher-order component of the Fourier transform disappears. Therefore, immediately after passing through the mask, the amplitude of light having a rectangular shape as shown in FIG. 8C and the distribution of the light intensity given by the square of the amplitude without losing the rectangularity on the image plane are shown in FIG. The distribution is as shown in d).

Next, a halftone type (attenuating type) phase shift mask will be described (for example, Japanese Patent Laid-Open No. 4-1).
No. 36854). This method is a method for improving the light intensity distribution at the pattern boundary portion by giving a slight transmittance to the portion that was originally the light shielding region and inverting the phase of the leaked light by 180 degrees. FIG. 9A is a plan view of a halftone phase shift mask.
(B) is sectional drawing in the AA line. This method is
As shown in FIGS. 9A and 9B, it is formed by providing a semitransparent film 3 made of chromium oxynitride (CrON) or the like on the transparent substrate 1. The transmissivity of this semitransparent film is generally set to about 3 to 15%, and the phase difference between the light passing through the transparent region and the semitransparent region is 180 degrees. As light travels through a material with a refractive index n, its wavelength becomes λ / n (where λ is the wavelength in a vacuum). Therefore, due to the optical path difference between air (n = 1) and the semitransparent film, A phase difference can be created. Here, the semitransparent film thickness t
Is set to t = λ / 2 (n−1) (n is the refractive index of the semitransparent film) so that the phase difference is 180 degrees.

FIG. 9C shows the amplitude of light transmitted through the transparent area and the semitransparent area. As shown in the figure, by controlling the phase of light leaking through the semitransparent film, light having different phases cancel each other at the edge portion of the semitransparent film 3, and as shown in FIG. 9D. , The light intensity distribution of the main pattern can be improved. However, in this halftone mask, light leaks over the entire surface of the mask. Therefore, when the semiconductor substrate is exposed using this photomask, this leaked light is generated at the boundary between adjacent exposure tilts. It has been pointed out that a pattern abnormality (film loss of the photosensitive resin) may occur due to overlapping a plurality of times.

Therefore, in Japanese Patent Laid-Open No. 6-282063, a light-shielding region frame (hereinafter referred to as a light-shielding band) is formed by a light-shielding film on a semi-transparent region around the exposure region, and the light-shielding band is used for exposure. A method for preventing film loss at the region boundary has been proposed. FIG. 10 shows an example of a halftone mask in which this light shielding band is formed. As shown in FIG. 10A, the light-shielding band is formed with a certain width along the outer circumference of the exposure amount area, and this is a semitransparent film 3 formed on the transparent substrate as shown in FIG. 10B. It has a structure in which the light shielding film 2 is placed on top.

[0015]

In the conventional modified illumination method described above, when applied to a repetitive pattern such as a line-and-space pattern, when the focus position is changed, the outer peripheral pattern becomes thin and disappears. There was a problem that caused it. This is because the modified illumination method has the effect of expanding the depth of focus only in the pattern having periodicity and has no effect in the isolated pattern, so that the depth of focus of the outer peripheral pattern in which the periodicity is disturbed decreases.

Such a problem is the same when a halftone phase shift mask is used in the modified illumination method. Although the depth of focus is increased in the pattern near the center of the periodic pattern, the outer peripheral pattern is thin and thin. There was a decrease. Similarly, for the isolated line pattern or space pattern, neither the modified illumination nor the halftone mask has sufficient depth-of-focus effect.
Therefore, the problem to be solved by the present invention is to provide a photomask having a sufficient depth-of-focus expansion effect even on the outer peripheral portion of a periodic pattern or an isolated pattern.

[0017]

The above-mentioned problems of the present invention are solved by forming an outermost peripheral portion of a periodic pattern or an isolated pattern apart from the periodic pattern as a semi-transparent pattern and another periodic pattern as a light-shielding pattern. can do.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION The photomask according to the present invention is one in which a mask pattern having periodicity is formed on a transparent substrate, and the pattern on the outer peripheral portion of the periodic pattern is a semi-transparent pattern and other patterns. Is a light-shielding pattern, and the semi-transparent pattern causes a predetermined phase difference (for example, 180 degrees) with respect to the light that does not pass through it.

Another photomask according to the present invention is
A mask pattern having periodicity and at least one isolated mask pattern formed on a transparent substrate, wherein the isolated mask pattern is a semi-transparent pattern, and the other patterns are light-shielding patterns, and the semi-transparent pattern Is to generate a predetermined phase difference with respect to light that does not pass through it.

In the photomask of the present invention, the pattern of the outermost peripheral portion of the periodic pattern has a halftone phase shift portion (having a transmittance of 2 to 20%, and the transmitted light and the transmitted light of the peripheral transparent portion). A phase difference of 180 degrees is generated), and the other pattern (internal periodic pattern) is used as a light-shielding pattern. The halftone pattern portion can reduce the light intensity as compared with the light-shielding pattern by canceling out lights having different phases by 180 degrees. Therefore, by making only the outer periphery of the periodic pattern half-tone, the dimension of the outer peripheral pattern is reduced and the film is reduced (erased)
Can be prevented.

Further, in another photomask of the present invention, the isolated pattern on the photomask is used as a halftone phase shift portion, and another periodic pattern is used as a light shielding pattern to reduce the light intensity of the isolated pattern. Then, by shifting the phase shift of the halftone phase shift unit from 180 degrees by a predetermined amount, the depth of focus range of that portion is shifted. That is, under modified illumination, the depth of focus itself does not expand even if the isolated pattern is a halftone, but by shifting the phase difference according to the step of the substrate to be exposed, a good pattern is formed above and below the step. can do.

[0022]

Next, embodiments of the present invention will be described with reference to the drawings. [First Embodiment] FIGS. 1A and 1B are a plan view and a sectional view, respectively, of a first embodiment of the present invention. Here, as the exposure apparatus, the reduction ratio is 1/5 (mask pattern size: pattern size on image plane = 5: 1), NA = 0.5.
5, annular illumination (σ = 0.8, 80% radius ratio shading), Kr
A reduction projection exposure apparatus of F excimer laser (wavelength λ = 248 nm) is used. The target pattern is a line-and-space pattern (11 lines) having a line width of 0.18 μm. In the following, all the dimensions of the pattern shape (excluding the thickness of the light shielding film, etc.) are shown on the semiconductor substrate. That is, it is five times that on an actual mask.

As shown in FIG. 1A, of the 11 line patterns (light-shielding patterns), the left and right outermost two lines are halftone patterns (transmittance 4%). In the mask structure, as shown in FIG. 1B, a semitransparent film 3 made of molybdenum oxynitride silicide (MoSiON) having a film thickness of 120 nm is formed on a transparent substrate 1 made of quartz.
Further, on the semi-transparent film 3, a 50 nm-thick chromium light-shielding film 2 is formed. Here, the phase difference caused by the semitransparent film 3 is 180 degrees.

Next, the effect of this photomask will be described in comparison with a normal mask. First, FIG. 2 shows a simulation result of the light intensity distribution of a normal mask (all line patterns are light shielding patterns). The horizontal axis is the substrate (image plane)
The upper left and right positions in FIG. 1 are shown, and the center line center position is 0.0 μm. The vertical axis represents the relative light intensity normalized with the light intensity in a sufficiently wide transparent region as 1. In FIG. 2, the focal position is changed from 0.0 to 1.0 μm in steps of 0.2 μm, and symbols a to f in the figure correspond to the focal positions of 0.0 μm to 1.0 μm, respectively. What is noticeable in this figure is that the focus position changes
In particular, the light intensity sharply increases in the line pattern portion on the outer peripheral portion. This indicates that when exposure is performed on a substrate coated with a photosensitive resin, a slight variation in the focal position causes a film loss or disappearance of the pattern in the outer peripheral portion.

FIG. 3 shows this ordinary mask and the photo of the present invention.
The simulation results of the mask are shown for comparison. here
Shows only the outermost peripheral pattern for easy comparison.
(The range of 1 to 2 μm on the horizontal axis of FIG. 2). Also here
The following assumptions are used when considering the film loss of a pattern. You
That is, the light intensity distribution at the focal position of 0.0 μm
The design dimension of the pattern is a slice level
(Light intensity at the pattern edge position) I₀ Based on
If the focus is changed, Imin / I ₀ There is
It is assumed that the pattern disappears when the value exceeds the value.
Imin / I when the pattern disappears₀ The value of is for the photosensitive resin
Although it depends, it is generally in the range of 0.6 to 0.8.
Therefore, here Imin / I₀ > 2/3 pattern erase
It is a condition of destruction. Main exposure condition I₀ Is 0.3
Then, in FIG. 3, when Imin is 0.2 or more, the pattern disappears.
Will be destroyed.

In the conventional mask, Imin exceeds 0.2 and the pattern disappears when the change in the focal position is 0.4 μm or more. That is, even if the modified illumination method is used,
It can be seen that the depth of focus does not increase as a whole due to the film loss (elimination) of the peripheral pattern. On the other hand, in the photomask of this embodiment, the focal position is 0.8 μm, which exceeds 0.2, but it can be seen that the dark portion is stably formed in the outermost layer portion as compared with the conventional mask. . Therefore, as for the depth of focus, the conventional mask has a depth of ± 0.2 μm, while the photomask of this embodiment has a triple depth of ± 0.6 μm.
It has been improved to μm. Although the KrF lithographic transmission type mask has been described in the above embodiments, the exposure light is not limited to the KrF excimer laser light in the above example, and other exposure light such as g and i rays of a mercury lamp or X rays may be used. It may be the wavelength.

[Second Embodiment] Next, a second embodiment of the present invention will be described with reference to the drawings. FIG. 4 (a),
(B) is a plan view and a sectional view showing a second embodiment of the present invention. Here, the reduction ratio is 1/5, NA = 0.
6. An i-line exposure device with annular illumination (σ = 0.7, 60% shading at the center) is used. As shown in FIG. 4, on the transparent substrate 1, 0.3 μm line and space patterns and isolated line patterns are mixed. Here, the line-and-space pattern is the light-shielding pattern 12, while the isolated line pattern is the halftone pattern 13. Here, the halftone pattern 13
Has a transmittance of 4% and a phase difference of 240 degrees (here, the thickness t of the semitransparent film is t = λ / 2 (n
The phase difference is 180 degrees when the relationship of -1) is satisfied, the phase difference is 180 degrees or more when it is thicker, and the phase difference is 180 degrees or less when it is thinner than that).

FIG. 5 shows the focus characteristics of the present mask and the conventional mask (the conventional mask in which the isolated line is a light-shielding pattern). In addition, here, the focus position "+/-"
In the direction of, the direction in which the substrate coated with the photosensitive resin approaches the lens is "+", and the direction in which the substrate is further away is "-". As is clear from FIG. 5, the focus characteristic of the line and space is flat and the depth of focus is 2.0 μm.
(−1.0 μm to +1.0 μm) is sufficiently obtained. On the other hand, the conventional isolated line exhibits a focus characteristic that is convex upward and has a depth of focus of 1.2 μm (-0.6 to +0.
6 μm), which is narrower than that of the line-and-space pattern [here, the dimensional variation of the depth of focus is within 10% of the target value (0.3 μm) (0.27 μm to 0.33).
(μm) is the range of the focus position).

Next, looking at the focus characteristics of the isolated line (0.3 μm width) of the photomask of the present invention, it can be seen that the pattern dimension becomes thicker than the conventional mask and the depth of focus shifts to the “−” side. I understand. That is, by making the isolated pattern a halftone, it is possible to prevent the pattern from becoming thin and, at the same time, to shift the focus position by shifting the phase difference from 180 degrees.
Therefore, when exposing the photosensitive resin 5 applied on the semiconductor substrate 4 having a step as shown in FIG. 6, when the isolated pattern is exposed under the step, the depth of focus of the isolated pattern is set to " By shifting to the "" side, it is possible to form a good pattern above and below the step.

As described above, in the photomask of the present invention, the isolated pattern having a narrow depth of focus is made into a halftone, and the phase difference is shifted from 180 degrees to shift the range of the depth of focus, thereby making a difference in level of the semiconductor substrate. A good pattern can be formed on the top and bottom. Here, the shift amount in the range of the depth of focus is controlled by the transmittance and the phase difference of the halftone portion, the transmittance is high, and the phase difference is 180 degrees.
The larger the deviation, the larger the shift amount. Further, the direction of the shift is shifted to the "-" side when the phase difference is larger than 180 degrees (thickness of the semitransparent film becomes thicker), and conversely becomes "+" when it becomes smaller (thickness becomes thinner). "Shift to the side. The exposure light in the second embodiment is not limited to the i-line.

[0031]

As described above, in the photomask of the present invention, only the outer peripheral portion of the periodic pattern and the isolated pattern are used as the halftone phase shift portion. Therefore, the film thickness of the photosensitive resin at that portion is reduced. And the disappearance of the pattern can be prevented. Further, particularly by controlling the phase shift due to the halftone, the optimum focus position of the halftone pattern can be shifted according to the step of the substrate, and a good pattern can be formed above and below the step.

[Brief description of drawings]

FIG. 1 is a plan view and a cross-sectional view showing a photomask of a first embodiment of the present invention.

FIG. 2 is a graph showing a light intensity cloth simulation result of a line and space pattern related to a conventional example.

FIG. 3 is a graph showing a light intensity distribution simulation result comparing the photomask of the first embodiment of the present invention with a conventional mask.

FIG. 4 is a plan view and a sectional view showing a photomask of a second embodiment of the present invention.

FIG. 5 is a graph showing simulation results of focus characteristics in the second embodiment of the present invention and the conventional photomask.

FIG. 6 is an explanatory diagram of an exposure state using the photomask according to the second embodiment of the present invention.

FIG. 7 is an explanatory diagram of a modified illumination method, which is one of conventional super-resolution techniques.

FIG. 8 is an explanatory diagram of a conventional photomask.

FIG. 9 is an explanatory diagram of a conventional halftone phase shift mask.

FIG. 10 is an explanatory diagram of a light-shielding band of a conventional halftone phase shift mask.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Transparent substrate 2 Light-shielding film 3 Semi-transparent film 4 Semiconductor substrate 5 Photosensitive resin 6 Photomask 12 Light-shielding pattern 13 Halftone pattern 101 Fly-eye lens 102a, 102b Aperture 103 Projection lens

Claims (4)

[Claims] 1. A photomask in which a mask pattern having periodicity is formed on a transparent substrate, wherein the outer peripheral pattern of the periodic pattern is a semi-transparent pattern, and the other patterns are light-shielding patterns, and the semi-transparent pattern. Is a photomask that produces a predetermined phase difference with respect to light that passes through it and light that does not pass through it. 2. The photomask according to claim 1, wherein the predetermined phase difference is 180 degrees. 3. A photomask in which a mask pattern having periodicity and an isolated mask pattern are formed on a transparent substrate, wherein the isolated mask pattern is a semi-transparent pattern, and the other patterns are light-shielding patterns, and the semi-transparent pattern. Is a photomask that produces a predetermined phase difference with respect to light that passes through it and light that does not pass through it. 4. The photomask according to claim 3, wherein the predetermined phase difference is an angle other than 180 degrees.