JP2666161B2 - Fine pattern projection exposure method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a so-called projection exposure method for forming a fine pattern such as an LSI on a substrate by using a projection lens.

[0002]

2. Description of the Related Art Conventionally, a projection exposure apparatus for forming a fine pattern such as an LSI has been required to have a high resolution. Therefore, a projection lens of a recent projection exposure apparatus has a resolution close to a theoretical limit determined by the wavelength of light.
Nevertheless, in order to respond to recent miniaturization of LSI patterns, higher resolution is required. There is a phase shift mask method as a method for answering this. In this phase shift mask method, a phase shift film is added at intervals of each transmission pattern so that a phase difference of π appears in transmitted light between adjacent patterns on the mask placed on the object plane of the projection lens. I do. The fact that the resolution is improved by this will be described with a simple example.

FIG. 17 is an explanatory diagram of a resolution limit of a mask by the phase shift mask method, and FIG. 18 is an explanatory diagram of a resolution limit of the mask by a normal irradiation method. Consider a mask in which many line figures having the same width are arranged in parallel at intervals equal to the line width. The mask M shown in FIG. 18 is a cross-sectional view taken along a plane perpendicular to the line direction, and the repetition period length of the line pattern of the mask M is d. In the phase shift mask MP shown in FIG. 17, a phase shift film S is provided for each interval of a transmission part so that a phase difference π is generated between light passing through adjacent transmission parts. That is, since the amplitude becomes negative at each interval of the pattern, the period becomes 2d and the DC component becomes zero. For this reason, the irradiation light I ₀ having the wave number k ₀ which is perpendicularly incident on the phase shift mask MP generates a wave I ₁ having an inclination in the optical axis z due to diffraction by the mask. The wave number of this wave is k ₁ = k ₀ s, where α is the diffraction angle.
In (α ′), k ₁ = 2π / 2d because the length of the pattern basic period is 2d. The electric field amplitude u of the transmitted light has a component in the ± x direction and is expressed by the following equation.

[0004]

(Equation 1)

[0005] Here, x is a coordinate in a pattern repetition direction. It is assumed that when the wave of Expression (1) goes to the projection lens, it passes around the aperture A (entrance pupil). That is, it is assumed that a finer pattern is not resolved by the projection lens. The light intensity on the image plane is 2 of the absolute value of the electric field amplitude.
Since it is proportional to the power,

| U | ² = (１ ／) | u ₀ | ² (1 + cos2k ₁ x) (2)

Thus, a pattern having a fundamental period d and a wave number k ₁ is reproduced. On the other hand, when a normal mask M without a phase shift mask is irradiated vertically with light I ₀ having a wave number k ₀ , FIG.
2k ₁ = 2k ₀ sin (α ′) ≒ k ₀ si
A wave I ₁ ′ having a wave number of n (2α ′) occurs, but this wave does not reach the image plane because it is blocked by the aperture A. That is, the pattern is not resolved by normal mask vertical irradiation.
From the above, it can be seen that the phase shift mask method improves the resolution.

[0008]

As described above, the phase shift mask method is effective for adjacent linear patterns, but is not necessarily effective for adjacent patterns having different sizes or isolated patterns. In addition, the step of adding a phase shift film in mask production causes a reduction in yield and an increase in mask production cost. Another method for responding to high resolution to cope with miniaturization is a ring light source aperture method. This method is disclosed in Japanese Patent Application No. 59-21 related to the same applicant.
This method uses a ring aperture as a light source aperture, as described in “Projection Exposure Apparatus” of No. 1269. In such a method, the grounds for its operation are based on experimental facts, and it is stated that in a ring light source aperture, the more the outer light source is used as much as possible, the higher the resolution becomes. However, there is no description on how to determine the size of the ring light source aperture for obtaining the highest resolution of the projection lens.

The present invention has been made in view of the above points, and an object of the present invention is to provide a phase shift mask method for a linear pattern without using a phase shift film for a mask pattern having an arbitrary shape and arrangement. It is another object of the present invention to provide a projection exposure method for giving the same resolution to that of an image forming system using a projection lens.

[0010]

The projection exposure method of the present invention in order to achieve the above object, according to an aspect of the (numerical aperture of the projection lens) sin .alpha '= with respect to the optical axis in the irradiation light for irradiating a mask /
The inclination of the angle α ′ defined by (reduction magnification of the projection lens) is given, and the value of the inclination is set corresponding to the basic various amounts of the projection lens.

[0011]

The operation of the present invention is first described by a linear pattern (wave number 2).
k ₁ ) will be described. Here, the mask M in FIG. 1 shows a cross section as viewed in a plane perpendicular to the line direction, and the repetition period length of the line figure is d. As shown in FIG. 1, a spatially coherent wave number k ₀ at an inclination angle α ′ to the optical axis z with respect to the normal mask M, that is, in the direction from the center of the optical axis z on the mask M to the helicopter of the aperture A. When the light I is irradiated, the wave I diffracted by the mask M and transmitted therethrough
_{Since 1} is a wave having an angle of 2α ′ with respect to incident light, it is a wave having a wave number k ₁ . Waves passing through the edge of the aperture A by straight in the incident direction is considered wave wavenumber k _1. Alternatively, assuming that the wave traveling straight in the incident direction is ₁ , the wave number of the diffracted wave is 2k1. Therefore, the electric field amplitude u of the transmitted light is

[0012]

(Equation 2)

## EQU1 ## The light intensity at the image plane is

[0014] ^{| u | 2 = (| u} 0 | 2/2) (1 + cos2k 1 x) ...... (4)

## EQU1 ## which is consistent with equation (2). That is,
It can be seen that when the normal mask M is irradiated from an oblique direction, an operation equivalent to that of the phase shift mask occurs. In addition, this is obtained by obliquely irradiating each transmission pattern with 0, π, 0,
It can be understood from the fact that a phase difference of π,. As described above, the fact that the phase shift method and the oblique irradiation method are equivalent in the case of the highest resolution was first discovered by the present inventors, and the present invention is based on this.

Next, the operation of the present invention when the periodicity of the mask pattern is in two directions will be described. Almost all integrated circuit patterns are composed of vertical and horizontal patterns. In this case, the same effect as a linear pattern can be expected by irradiating obliquely in the x and y directions. The operation when irradiation in two directions or irradiation in four directions with further improved symmetry will be described with reference to FIG.

Using incoherent illumination from multiple directions is equivalent to irradiating independently of each other,
No extra interference occurs. In FIG. 2A, the horizontal axis represents the inclination amount βx in the x direction (wave number βx), and the vertical axis represents the inclination amount βy in the y direction.
Taking (wave number βy), points AP1 to AP4 indicating the amount of inclination of the light source are placed at four symmetrical points on the circle A1 about the origin. The radius R of A1 is set equal to the numerical aperture NA1 when viewing the exit pupil of the projection lens from the center of the optical axis of the image plane. Here, the irradiation angle corresponding to the numerical aperture of the projection (optical) lens indicates α ′ in the following equation with respect to a reduction ratio of 1 / m.

Sinα ′ = (1 / m) sinα (5)

Where sin α is the numerical aperture of the projection lens. For example, when the reduction ratio is 1/5 and the numerical aperture of the projection lens is 0.5, sinα '= (1/5) × 0.5, and
α ′ ≒ 5.7 °. Since the gist of the present invention is to perform oblique irradiation with a gradient equal to this value, a point indicating the irradiation light comes on A1. Circle A1 ', A1 "is a mask pattern when decomposed into spatial harmonics, .A1 high order component corresponding to the wave number k _1, 2k ₁ respectively indicate the extent to which incorporated into the projection lens' that if the optical axis The position of the best resolution circle that would be obtained if illuminated in parallel, A1 "indicates the range in which the resolved wavenumber component obtained by the present invention is taken, and the center Px of the parallel illumination indicates the range. Value.

The optical transfer function OTF of FIG.
As shown in ( ₁₎ , the maximum wave number in the x-axis and y-axis directions is twice the value of k ₁ in the case of parallel irradiation with the optical axis, that is, 2k ₁ . In the middle direction between the x direction and the y direction, the route of the optical axis parallel irradiation is doubled, which is suitable for the pattern to be resolved. Next, if it is required that the mask pattern be symmetrical about the optical axis, the irradiation light will have a numerical aperture N with respect to the optical axis.
Irradiation light having an inclination corresponding to A1 may be irradiated isotropically (axially symmetric). At that time, as shown in FIG. 3A, the shape becomes a fine annular shape, and the optical transfer function OTF is

F (k) = (1 / π) cos ⁻¹ (k / 2k ₁ ) (6)

## EQU1 ## This is shown in FIG. Compared with a normal irradiation method including a component from a component parallel to the optical axis to a certain inclination, the characteristics slightly decrease at high frequencies, but maintain a considerably large value up to the highest frequency, which is superior to the normal irradiation method. If the numerical aperture of the irradiation light does not correspond properly to the numerical aperture of the projection lens system, there is no great difference from the case of normal partial coherent irradiation even if an annular irradiation system is used. It is understood from the calculation.

Next, a method for further improving the imaging characteristics and improving the contrast will be described. In the OTF shown in FIG. 3B, the component at k = 0 is sharply large, which means that the DC component of the spatial frequency component of the pattern is large, and lowers the contrast. To remove the DC component, the DC component of the tilt irradiation may be removed when passing around the projection lens opening. However, since the peripheral part also passes through the higher-order components of the pattern, it is necessary to set the transmittance of the peripheral part in consideration of the resolution and the contrast in accordance with the oblique irradiation. By using such a projection lens aperture, more excellent imaging characteristics can be obtained. Here, the projection lens numerical aperture refers to a portion of an aperture stop attached to a position where an image of a light source is formed in a projection lens system. O when the amplitude transmittance is 50% only on the arc line around the projection lens aperture
TF is shown in FIG.

In the case of the four-point oblique irradiation method shown in FIG.
In order to remove the sharp DC component of the OTF of (b), a projection lens aperture as shown in FIG. 5 is used. Where AP
1 ', AP2', AP3 ', AP4' are portions corresponding to oblique irradiation points, and their amplitude transmittances are set to "1/2". The amplitude transmittance in the square area indicated by AP5 is set to “1”, and the four half-moon-shaped areas indicated by AP6 do not transmit light, that is, the transmittance is set to “0”. By using such a projection lens aperture, an effect of improving contrast can be obtained.

To supplement the OTF used in the above description, in the case of a partially coherent light source, the mutual transmission coefficient must be used, which is a function of two wave numbers and depends on the type of pattern. I do. Therefore, the OTF is defined by a mutual transmission coefficient that has been made into one variable excluding the pattern dependency, and is used for explanation. FIG. 6 shows a list of the relationship between the arrangement of the light sources and the transmittance of the opening of the projection lens. Here, FIG. 6A shows the case of one-point oblique irradiation, and FIG.
FIG. 6C shows the case of point oblique irradiation, FIG. 6C shows the case of four-point oblique irradiation shown in FIG. 5, and FIG. 6D shows the case of using an annular irradiation system. 6 (a1) to 6 (d1) show the transmittance of the projection lens aperture corresponding to the arrangement of each light source in FIGS. 6 (a) to 6 (d). That is, FIG. 6A1 shows a substantially circular area AP7.
Is a projection lens opening in which the transmittance of “1” is set to “1”, and the transmittance of the arc-shaped region AP8 corresponding to the oblique irradiation point is set to “１ ／”. FIG. 6 (b1) shows that the transmittance of the portions AP1 ′ and AP2 ′ corresponding to the oblique irradiation points is “「 ”, the transmittance of the substantially rectangular area AP9 is“ 1 ”, and three half-moon areas. AP
6 shows a projection lens aperture with the transmittance of “0” being “0”, and FIG.
(c1) corresponds to FIG. 5 described above, and FIG. 6 (d1) illustrates a circular area AP1.
A projection lens aperture is shown in which the transmittance of 0 is “1” and the transmittance of the annular area AP11 is “１ ／”.

The above description is for clarifying the basic principle. Here, the outline of the optical system of an actual reduction projection exposure apparatus will be supplementarily described. FIG. 7 shows the basics of an optical system that forms an image on a wafer surface 15 by irradiating a mask pattern (reticle) 11 with illumination light from a point light source being slightly converged using a condenser lens (not shown). The configuration is shown. Projection lens systems are classified into a first lens system 12 and a second lens system 13 in terms of function. Although each lens is represented by one lens in FIG. 7, it is actually composed of a plurality of lenses for optimizing aberrations and the like, but from the functional aspect, each lens can be represented by one lens. An aperture stop 14 is provided between the first lens system 12 and the second lens system 13, and an image of the illumination point light source is usually formed at the center thereof. The image (virtual image) of the aperture stop 14 viewed through the first lens system 12 is called an entrance pupil. Mask pattern 1
It can be seen that the light diffracted by 1 (from p, p ') can pass through the projection lens system and contribute to image formation on the wafer surface 15 within a range of angles where the entrance pupil 16 can be seen. However, the exit pupil is the upper ∞.

Aperture A of FIG. 1 for illustrating the basic principle
Means that actually means the entrance pupil described above. In the image forming part, a telecentric optical system in which a principal ray passing through the center of the aperture stop 14 is perpendicular to the wafer surface 15 is usually used. Therefore, the focal length f ₂ of the second lens system 13 is as shown in FIG. The aperture stop 14 and the wafer surface 15 are set at positions. In the case of the reduction magnification is 1/5, the mask pattern (reticle) 11 installed at a position of the focal length f ₁ of the first lens system 12, and f ₁ / f ₂ ≒ 5.

FIG. 7 shows a point p on the optical axis of the mask pattern 11.
And the diffracted light from the off-axis point p '
A state where an image is formed on Q 'is shown. Each light beam is a parallel light between the first lens system 12 and the second lens system 13. Focusing on the aperture stop 14, the diffracted light having the largest diffraction angle passes around the aperture stop 14 without depending on the positions of p and p 'on the mask pattern 11. That is, the aperture stop 14 corresponds to the Fourier transform surface of the mask pattern 11. In order to improve the image forming characteristics by adjusting the transmittance of the projection lens opening, if a filter is provided at the aperture stop 14 and the light transmittance is adjusted, the entire surface of the mask pattern 11 can be adjusted. It is understood that the imaging characteristics (contrast) are improved over the entire range.

[0029]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be shown below with specific numerical values. FIG. 8 shows an example of the mask pattern, and the pattern 1
The rectangular portion shown by the pattern 3 is a light transmitting portion, and the other region does not transmit light. The entire area of 3 μm × 3 μm is x
It is assumed that they are arranged infinitely periodically in the direction and the y direction. The width a of each pattern is 0.3 μm, and the distance b between the patterns is also 0.3 μm.

FIG. 9 shows that the range of the tilt irradiation is NAs = 0.3.
7 to 0.40, and the numerical aperture NA1 of the projection lens is 0.4
0 is a numerical example of the present invention when the light wavelength λ is 0.365 μm and the reduction magnification is 1/1. FIG. 9A is a contour line of the light intensity distribution in a 3 μm × 3 μm area on the image plane. The numerical values of the maximum intensity value max and minimum value min are displayed, and these are max = 3.830900e-02, min = 8.079000e, respectively.
-04. Contour lines are drawn as bold lines when the 0-max value is divided into ten parts. Distance from the coordinate origin to the center of the display frame, to the end in the x direction of the frame alx and distance in the y direction
aly is expressed in μm. FIG. 9B is a cross-sectional view of FIG.
FIG. 9 is a diagram showing the light intensity distribution along a line in the x direction with a mark, and patterns 1 to 3 correspond to the patterns indicated by reference numerals 1 to 3 in FIG. Numerical values shown as (xa, ya) and (xb, yb) represent coordinate values of both ends of a line segment marked with a cross in FIG. 9A in μm units, and (xa, ya) = -1.500
542,0.00674, (xb, yb) = 1.493813, 0.026966.

As can be seen from FIG. 9, the light intensity decreases in the order of pattern 3, pattern 2, and pattern 1. When the resist photosensitive film applied on the substrate is exposed by this light intensity distribution, whether or not the pattern 1 is resolved after development depends on the resist system, but it is estimated that the situation is close to the limit. FIG. 10 shows the light intensity distribution of an image when the pattern 1 and the pattern 3 are left as they are in the mask of FIG. 8 and a phase shift film for shifting the phase by π is added to the pattern 2.
The imaging conditions are the same as in FIG. 9 except that the irradiation conditions are NAs = 0.0 to 0.2. As can be seen from FIG. 10 (a), pattern 1 and pattern 3 are connected,
Not resolved. This is because pattern 1 and pattern 3 are in phase. The phase shift mask method is an example in which the phase arrangement of 0-π is difficult depending on the pattern figure.

FIG. 11 shows a light intensity distribution when the mask of FIG. 8 without phase shift is imaged under the same optical conditions as in FIG. Patterns 1, 2, and 3 are not resolved.
FIG. 12 shows light formed by the oblique irradiation method of the present invention using a mask in which the patterns 1, 2, and 3 in the mask of FIG. 3 shows an intensity distribution. The imaging conditions are the same as in FIG. FIG. 13 shows a case where the same mask as in FIG. 12 is formed under normal conditions, that is, under the same conditions as in FIG. As can be seen by comparing FIGS. 12 and 13, it is clear that the present invention is effective. Although an example of a phase shift mask is not shown for the mask in this case, it is because design of this phase shift arrangement is considerably difficult.

FIG. 14 shows a light intensity distribution when an image is formed by the four-point oblique irradiation method of the present invention using the mask of FIG. Here, a square area is set in AP1 of FIG. 2A (βx = 0.35 to 0.4, βy = 0.55 to 0.05),
Similar areas were set for AP2, AP3, and AP4. As is clear from FIG. 14, the present invention is more effective than the conventional method. FIG. 15 shows an image formed by the four-point oblique irradiation method of the present invention using an inversion mask so that the patterns 1, 2, and 3 do not transmit light in the mask of FIG. 3 shows the light intensity distribution of the sample. As can be seen from FIG. 15, it is clear that the present invention is more effective than the conventional method.

Next, numerical examples in which the transmittance of the opening of the projection lens is adjusted to improve the contrast will be described. FIG. 16 shows a case where the transmittance at the portion of NA1 = 0.37 to 0.40 around the opening of the projection lens is set to 50% under the same inversion mask as in FIG. , NA1 = 0.0 to 0.37 are light intensity distribution diagrams when 100% transmittance is set. It can be seen that the contrast is approximately doubled as compared with FIG.

Although the calculation results using the i-line of the Hg lamp as a light source have been specifically described above, similar effects can be obtained when the g-line of an Hg lamp or various excimer lasers are used as a light source. Is clear. Further, since the present invention is based on a method in which the periphery of the projection lens is slightly emphasized, it is expected that the influence of the aberration of the lens or the depth of focus will be affected by the conventional method. Using an aperture stop slightly lower than the maximum numerical aperture of the lens,
If the present invention is applied with a slightly reduced numerical aperture, the frequency band can be sufficiently improved, the depth of focus is sufficient, and projection exposure with less aberration can be performed.

[0036]

As described above, according to the present invention, an image is formed at the highest resolution of the projection lens by setting the range of the inclination of the irradiation light used in the conventional apparatus according to the numerical aperture of the projection lens. be able to. In addition, there is an advantage that resolution equivalent to that of the phase shift mask method using the conventional apparatus can be achieved for any pattern without using a phase shift mask. As described above, according to the present invention, there is an effect that the integration degree and the reliability can be improved in forming a fine pattern such as an LSI.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is an explanatory diagram of the resolution of a four-point or two-point oblique irradiation method according to one embodiment of the present invention.

FIG. 3 is an explanatory diagram of the resolution of an axially symmetric oblique irradiation method according to another embodiment of the present invention.

FIG. 4 is a diagram showing an OTF when an amplitude transmittance on an arc line around an opening of a projection lens according to the present invention is 50%.

FIG. 5 is a diagram showing a projection lens aperture for improving contrast in the four-point oblique irradiation method according to the present invention.

FIG. 6 shows the relationship between the arrangement of the light source and the transmittance of the opening of the projection lens according to the present invention, wherein (a) to (d) show the arrangement of the light source, and (a1) to (d1). FIG. 3 is a diagram showing a projection lens aperture corresponding to each light source of (a) to (d).

FIG. 7 is a schematic diagram of an optical system for forming an image on a wafer surface by irradiating irradiation light from a point light source to a mask pattern via a condenser lens according to the present invention.

FIG. 8 is a diagram illustrating an example of a mask pattern used in the present embodiment.

FIG. 9 is a light intensity distribution diagram of an image obtained by the tilt irradiation method of the present invention.

FIG. 10 is a light intensity distribution diagram of an image formed by a conventional phase shift mask.

FIG. 11 is a light intensity distribution diagram of an image obtained by a normal irradiation method.

FIG. 12 is a light intensity distribution diagram of an image of the inversion mask according to the present invention.

FIG. 13 is a light intensity distribution diagram of an image of a reversal mask by a normal method.

FIG. 14 is a light intensity distribution diagram of an image by irradiation light by the four-point oblique irradiation method of the present invention.

FIG. 15 is a light intensity distribution diagram of an image of a reversal mask with four-point oblique irradiation light of the present invention.

FIG. 16 is a light intensity distribution diagram of an image when the transmittance of the reversal mask around the opening of the projection lens according to the present invention is 50%.

FIG. 17 is an explanatory diagram of a conventional phase shift mask method.

FIG. 18 is an explanatory diagram of a normal irradiation method.

[Explanation of symbols]

Inclination angle with respect to the optical axis-frequency M normal mask z optical axis α with wavenumber of 'the irradiated light I ₁ k of A aperture I wavenumber k _{0 0} sin .alpha

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Emi Tamechika 1-6-1, Uchisaiwai-cho, Chiyoda-ku, Tokyo Nippon Telegraph and Telephone Corporation (72) Katsuyuki Harada 1-6-1, Uchisaiwai-cho, Chiyoda-ku, Tokyo Nippon Telegraph and Telephone Corporation (56) References JP-A-61-91662 (JP, A)

Claims (5)

(57) [Claims] 1. For a light beam illuminating an object mask on which a pattern is drawn, sin α ′ = (numerical aperture of projection lens) /
Fine pattern projection exposure method characterized by providing the inclination angle alpha 'as defined in (reduction magnification of the projection lens) and the optical axis. 2. The fine pattern projection exposure method according to claim 1, wherein a plurality of mutually incoherent irradiation lights satisfying the above requirements are superimposed. 3. A fine pattern projection exposure method according to claim 2, wherein oblique irradiation is performed from four equal angles or two directions perpendicular to each other when viewed in a plane perpendicular to the optical axis. . 4. The fine pattern projection exposure method according to claim 2, wherein oblique irradiation is performed at an angle symmetrical with respect to the optical axis and at an angle corresponding to the numerical aperture of the projection lens. 5. A projection lens opening according to claim 1, wherein the projection lens opening has a transmittance adjusted around the opening when the oblique irradiation range is made to correspond to the projection lens opening. Pattern exposure method.