JPH10242018A - Exposure optical system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an exposure optical system with superior parallelism in spite of low cost and a short optical path length. SOLUTION: The optical system consists of a light source 10, a condensing means 20, a secondary light source-forming means 30, a collimator 40 and a mirror 50. In this case, when an angle of incidence of a light beam A to the mirror A is θ and a parallel error of a light beam B and the light beam A is ϕ, the mirror 50 is elastically deformed so that an angle of incidence of the light beam B to the mirror 50 becomes θ+ϕ/2. The light beam A is a principal ray beam on an optical axis to enter the mirror 50, and the light beam B is a principal ray beam passing outside the optical axis to enter the mirror 50.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for exposing a pattern drawn on a mask to a semiconductor substrate, a glass substrate, or the like (hereinafter, referred to as a substrate), and more particularly, to exposing a pattern with a small space between the mask and the substrate. The present invention relates to an optical system of a so-called proximity exposure apparatus.

[0002]

2. Description of the Related Art The optical system of a conventional proximity exposure apparatus has a structure as shown in FIG. That is, the lamp 1
The light emitted from 1 is collected by the elliptical mirror 21 to the secondary light source creating means 30, where the secondary light source is created. Light emitted from the secondary light source is converted into parallel light by the collimator 40, and irradiates the mask and the substrate provided on the exposure surface 60.

[0003]

However, if the performance of the collimator 40 is insufficient and there is an error in the parallelism of the light beam, the irradiation angle of the light beam will differ depending on the location on the mask. For example, in FIG. 2, when the proximity gap 65 is 500 microns and there is an error of 0.1 degree in the parallelism of light rays, the transfer position shift 66 of the pattern 62 drawn on the mask 61 becomes 0.9 microns. ,
It hinders the production of semiconductors and liquid crystals. For this reason, conventionally, the following two approaches have been taken to improve the performance of the collimator 40.

[0004] The first approach is to use a parabolic mirror for the collimator. It is widely known that the use of an off-axis parabolic mirror as shown in FIG. 3 allows the light emitted from one point to be converted into perfect parallel light. However, in order to manufacture an off-axis paraboloid, it is necessary to generate a paraboloid having a diameter D including the axis as shown in FIG. 3, and it is not easy to polish because the curvature varies depending on the position of the mirror. . For this reason, the use of an off-axis parabolic mirror for the collimator has the disadvantage of extremely high costs.

[0005] A second approach is to use a spherical mirror for the collimator and make the focal length as long as possible. 1
If you try to convert light emitted from a point into parallel light with a spherical mirror,
Parallel errors occur due to spherical aberration. Here, the parallel error is the F number of the spherical mirror (the value obtained by dividing the focal length by the aperture).
Since it is inversely proportional to the square of, the parallel error can be reduced to a level that does not hinder practical use by increasing the focal length. Generally, if the F-number is set to 5 or more, the level becomes practically acceptable. However, increasing the focal length has the disadvantage of increasing the optical path length. In particular, in recent years, the area of the liquid crystal substrate has been increased, and the exposure size has to be about 1 m diagonally. In this case, a focal length of 5 m is required,
The optical path length from the secondary light source to the exposure surface is as long as 10 m. Such an increase in the size of the apparatus not only increases the cost of the apparatus, but also increases the area occupied by an expensive clean room. Therefore, an object of the present invention is to provide an exposure optical system in which the parallelism of exposure light is improved without increasing the cost of the apparatus and without increasing the size of the apparatus.

[0006]

A light source 10 and a condensing means 2 are provided.
0, secondary light source creating means 30, collimator 40,
In the exposure optical system including the mirror 50, when the incident angle of the light beam A to the mirror 50 is θ and the parallel error between the light beam B and the light beam A is φ, the incident angle of the light beam B to the mirror 50 is θ +
The mirror 50 is elastically deformed so as to be φ / 2. Here, the ray A is a principal ray on the optical axis that enters the mirror 50, and the ray B is a principal ray that passes off the optical axis and enters the mirror 50.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present embodiment is 740 mm × 620.
FIG. 1 shows an optical system for exposing a region of mm with ultraviolet light (main wavelength: 365 nm) having a uniform intensity. Reference numeral 10 denotes a light source. In this embodiment, an ultra-high pressure mercury lamp is used. However, a xenon mercury lamp or the like may be used depending on a photosensitive material applied on a substrate. Reference numeral 20 denotes a condensing unit, and an elliptical mirror is used in the present embodiment. In addition, the condensing means includes a spherical mirror, a condenser lens, and the like. Reference numeral 30 denotes a secondary light source creating means. In the present embodiment, it comprises a rod 31 and a lens 33.
The light condensed by the light condensing means 20 is input to the rod 31, and total reflection is repeated inside the rod 31, so that a uniform light intensity distribution is obtained on the emission end face 32. The lens 33 enlarges and projects the emission end surface 32 of the rod onto the exposure surface 60, thereby realizing exposure with a uniform illuminance distribution. The secondary light source creating means also includes a fly-eye lens as shown in FIG. This realizes a uniform illuminance distribution on the exposure surface 60 by dividing and superimposing a light beam.

[0008] A spherical mirror is used for the collimator 40. It is located 2.46m from the secondary light source,
Its focal length is 2.5m, aperture is 840mm x 720m
m. The off-axis angle is set at 28 degrees. In the present invention, since the parallelism of light rays is corrected by the mirror 50 as described later, it is not necessary to use an expensive off-axis parabolic mirror for the collimator, and it is not necessary to lengthen the focal length of the spherical mirror unnecessarily. Absent. However, of course, the collimator is not limited to a spherical mirror and includes other aspheric mirrors and lenses.

The mirror 50 is formed by evaporating aluminum on glass having a thickness of 5 mm, and has a size of 980 mm × 84 mm.
0 mm. This mirror is housed in a holder 51. FIG. 5 shows the parallelism of light rays on the exposure surface 60 when the mirror 50 is a perfect plane. The starting point of the arrow indicates the position on the exposure surface 60, and the direction of the arrow indicates the falling direction of the light beam. The length of the arrow indicates the parallelism of the light beam. Since the collimator 40 has spherical aberration, the parallelism has reached 0.42 degrees at the worst, and the exposure position error cannot be ignored. Next, FIG. 6 shows the parallelism of light rays when the mirror 50 is elastically deformed into the shape shown in FIG. The mirror 50 can be easily deformed because it is relatively thin. FIG. 7 is a contour diagram of the mirror 50, and it can be seen that the mirror 50 is deformed by about ± 2 mm. As shown in FIG. 6, the parallelism is 0.06 degrees at the worst value, which is improved to a level that does not hinder practical use. Here, the mirror 50
The method for obtaining the ideal shape of the above will be described with reference to FIG.
In the following description, a ray A is a principal ray on the optical axis that enters the mirror 50, and a ray B is a principal ray that passes off the optical axis and enters the mirror 50.

Here, the angle between the ray A and the normal N of the mirror at the point where the ray A enters the mirror 50 is defined as θ. Further, a light ray after the light ray A is reflected by the mirror 50 is defined as P. Ray P forms an angle θ with normal N, and ray A and normal N
Is in the plane to make. On the other hand, when the parallel error between the light beam B and the light beam A is φ, the light beam P
In order to be parallel, the mirror normal M at the point where the light beam B enters the mirror 50 as shown in FIG.

By repeating the above operation at many points of the mirror 50, the inclination of each point of the mirror 50 is determined. By integrating this over the entire surface, the ideal shape of the mirror 50 is determined.
On the other hand, if the ideal shape is created by polishing, the cost becomes extremely high. Therefore, in a preferred embodiment of the present invention, the above-described shape is obtained by elastically deforming the mirror 50. The amount of elastic deformation that occurs when the mirror 50 is placed in the holder 51 and pressure is applied to some places is slightly different from the ideal shape. For this reason,
After being reflected by the mirror 50, the luminous flux does not always become parallel, but as shown in FIG. 6, there is no practical problem if the residual amount is 0.06 degrees. In addition, as shown in FIG. 7, the deformation amount is about ± 2 mm, which is within the elastic deformation area of the mirror 50.

[0012]

According to the present invention, it is not necessary to make the collimator aspherical in order to improve the parallelism of the exposure light beam. Therefore, the cost of the exposure optical system can be reduced as compared with an exposure optical system using an aspherical mirror as a collimator.

Further, according to the present invention, it is not necessary to increase the focal length of the collimator to improve the parallelism of the exposure light beam. Therefore, the total optical path length of the exposure optical system can be shortened, and the floor area where the exposure apparatus is installed can be reduced.

[Brief description of the drawings]

FIG. 1 is an overall configuration diagram of an exposure optical system according to the present invention.

FIG. 2 is a diagram illustrating that an error in the degree of parallelism of exposure light causes transfer deviation of a pattern.

FIG. 3 is a diagram illustrating an off-axis parabolic mirror.

FIG. 4 is a diagram illustrating a fly-eye lens.

FIG. 5 is a diagram illustrating the parallelism of exposure light when no correction is performed by a mirror 50;

FIG. 6 is a diagram illustrating the parallelism of exposure light when correction has been performed by a mirror 50;

FIG. 7 is a contour diagram showing the amount of elastic deformation of the mirror 50;

FIG. 8 is a diagram illustrating how light rays A and B enter a mirror 50;

FIG. 9 is a diagram illustrating a conventional exposure optical system.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Light source 20 Condensing means 30 Secondary light source creation means 40 Collimator 50 Mirror

Claims (1)

[Claims] In an exposure optical system including a light source, a light condensing means, a secondary light source creating means, a collimator, and a mirror, an incident angle of a light beam on the mirror is defined as θ. When the parallel error between the ray B and the ray A is φ,
An exposure optical system, wherein the mirror 50 is elastically deformed so that the incident angle of the light beam B on the mirror 50 becomes θ + φ / 2. Here, the ray A is a principal ray on the optical axis that enters the mirror 50, and the ray B is a principal ray that passes off the optical axis and enters the mirror 50.