JP2003050455A - Mask and method of manufacturing the same, optical element and method of manufacturing the same, illumination optical device and exposure device having this optical element - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method of manufacturing a diffraction optical element which is capable of efficiently converting an incident luminous flux to a luminous flux having a prescribed sectional shape. SOLUTION: The prescribed luminous flux section is divided to obtain a plurality of first luminous flux sectional segment regions (S11). The total sum of the lengths of a plurality of the first luminous flux sectional segment regions is calculated (S12) and a prescribed unit length is obtained by equally dividing the total sum of the lengths (S13). A plurality of the second luminous flux sectional segment regions are obtained in accordance with the prescribed unit length (S14). The prescribed optical element is divided to a plurality of partial optical elements by using the coordinate data on the prescribed optical element corresponding to the coordinates of a plurality of the second luminous flux sectional segment regions (S15). The basic optical element is constituted by densely arranging a plurality of the partial optical elements (S16).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mask and a manufacturing method thereof, an optical element and a manufacturing method thereof, and an illumination optical apparatus and an exposure apparatus equipped with the optical element. In particular, the present invention relates to the manufacture of a diffractive optical element suitable for an illumination optical apparatus of an exposure apparatus for manufacturing a microdevice such as a semiconductor element, an image pickup element, a liquid crystal display element, a thin film magnetic head, etc. in a lithography process.

[0002]

2. Description of the Related Art For forming a circuit pattern of a semiconductor element or the like,
Generally, a process called photolithography technique is required. In this step, a method of transferring a reticle (mask) pattern onto a substrate such as a semiconductor wafer is usually adopted. A photosensitive photoresist is coated on the substrate, and the circuit pattern is transferred to the photoresist according to the irradiation light image, that is, according to the pattern shape of the transparent portion of the reticle pattern.

In a projection exposure apparatus, an image of a circuit pattern to be transferred drawn on a reticle is projected and exposed on a substrate (wafer) via a projection optical system. In this type of projection exposure apparatus, an optical integrator such as a fly-eye lens is used in the illumination optical system for illuminating the reticle, and the intensity distribution of the illumination light with which the reticle is irradiated is made uniform. The reason why the intensity distribution of the illumination light irradiated on the reticle is made uniform by using an optical integrator such as a fly-eye lens will be described below.

In a projection exposure apparatus using a fly-eye lens, a light beam emitted from a light source (eg, KrF excimer laser) is guided to a fly-eye lens via a beam expander optical system and a vibrating mirror. The light that has passed through the fly-eye lens passes through the aperture stop and then illuminates the reticle (mask) through the condenser lens. here,
The incident surface of each lens element that constitutes the fly-eye lens,
The reticle surface is optically conjugate with each other.

Therefore, as a result, the light flux incident on the fly-eye lens is divided into wavefronts for each lens element unit of the fly-eye lens, and is superimposed on the reticle surface. Therefore, even when the incident light flux to the fly-eye lens has a Gaussian distribution-like intensity distribution on the incident surface, the intensity distribution becomes uniform to some extent in the lens element unit of the fly-eye lens, and they overlap on the reticle surface. As a result, the illuminance distribution is extremely uniform on the reticle surface because it is averaged. Less than,
A system that uses one fly-eye lens is called a single fly-eye lens system.

A system in which two fly-eye lenses are used to repeat the process of wavefront division and superposition thereof twice, that is, a double fly-eye lens system is also known. In a projection exposure apparatus using a double fly-eye lens system, a light beam from a light source such as an excimer laser is converted into a desired light beam cross-sectional shape via an expander, and then a mirror and a quartz prism for relaxing the polarization of the light beam. Through a first fly-eye lens (secondary light source forming means) including a plurality of lens elements, and a large number of secondary light sources are formed on the exit surface or in the vicinity thereof.

Divergent light beams from a large number of secondary light sources are condensed through a relay lens and uniformly illuminate the incident surface of a second fly-eye lens (tertiary light source forming means) composed of a plurality of lens elements in a superimposed manner. . As a result, a large number of tertiary light sources corresponding to the product of the number of lens elements of the first fly-eye lens and the number of lens elements of the second fly-eye lens are formed on or near the exit surface of the second fly-eye lens. To be done. Reticles (masks) with a pattern drawn on them are projected and exposed after being condensed through a condenser lens group after the diameters of the light beams from a large number of tertiary light sources are limited by an aperture stop.
Are uniformly illuminated in a superimposed manner.

Here, a field stop for determining the illumination range on the reticle is arranged in the condenser lens group. Then, the pattern formed on the reticle (mask) is projected on the substrate (wafer) that is the object to be exposed through the projection optical system based on the uniformly illuminated illumination light. The advantages of the double fly-eye lens system over the single fly-eye lens system are as follows.

Regarding the effect of making the illuminance of the illumination light for illuminating the reticle uniform, the effect increases as the number of divisions of the fly-eye lens increases, but the manufacturing cost of the fly-eye lens is almost equal to the number of divisions of the fly-eye lens. Proportional. Therefore, if many wavefront divisions are to be realized with the single fly-eye lens system, the manufacturing cost of the fly-eye lens will be enormous. On the other hand, in the double fly-eye lens system, since the product of the number of divisions of the first fly-eye lens and the number of divisions of the second fly-eye lens is effectively the total number of divisions of the optical system, it is large for the fly-eye lens. There is an advantage that a high-performance illumination system can be obtained without incurring manufacturing costs.

Further, in the single fly-eye lens system, a light beam having a light intensity distribution depending on the light source is directly incident on the fly-eye lens. Therefore, if the light intensity distribution of the light flux changes due to the vibration of the light source or the like, the spatial coherency of the projection exposure apparatus changes. On the other hand, in the double fly-eye lens system, the light intensity distribution on the incident surface of the second fly-eye lens is subject to the homogenizing effect of the first fly-eye lens, and thus is affected by the vibration of the light source. It rarely changes. In other words, the double fly-eye lens system has an advantage that vibration of the light source or the like hardly affects the imaging performance.

Further, in the double fly-eye lens system, when the aperture stop arranged near the exit surface of the second fly-eye lens is replaced, the illuminance uniformity is broken, that is, the change amount from the ideal Koehler illumination state. There is also the advantage that there are fewer than in the case of a single fly eye lens system.

[0012]

In recent years, the performance such as resolution required for this type of exposure apparatus is approaching the theoretically calculated limit. As is well known,
The optimum optical system constants (numerical aperture of projection optical system, numerical aperture of illumination system, etc.) vary depending on the reticle pattern.
Since the exposure is performed in the vicinity of the theoretical limit of the apparatus capability, the apparatus side is naturally required to be able to select the optimum optical system constant according to the reticle pattern.

That is, at least the numerical aperture of the illumination system needs to be variable. In the above-mentioned double fly-eye lens system, in order to make the numerical aperture of the illumination system variable, the size of the aperture (light transmitting part) of the aperture stop arranged near the exit surface of the second fly-eye lens, that is, the aperture size, is set. It is common to make the aperture variable or switch the aperture stop. However, if the aperture diameter is changed or the aperture stop is simply switched, the loss of the light amount in the aperture stop increases and the illuminance on the reticle or the wafer decreases.

A decrease in illuminance in this type of exposure apparatus means a decrease in throughput and, as a result, the cost of manufactured products is increased. For this reason, "illuminance" is one of the important items among various specifications of the exposure apparatus, and it is necessary to avoid a decrease in illuminance as much as possible. In the double fly-eye lens system, multiple first fly-eye lenses with different focal lengths are used to reduce the illuminance.
A method has been proposed in which switching is performed with aperture stops having different aperture diameters. In this method, for example, when the aperture diameter is reduced, the first fly-eye lens having a long focal length is switched to.

In this case, since the luminous flux is concentrated near the center of the incident surface of the second fly-eye lens, the light amount is hardly lost even if the aperture diameter is small, and the illuminance is hardly reduced. As described above, when only the aperture diameter of the aperture stop changes, it is possible to suppress the decrease in illuminance by simply switching the plurality of first fly-eye lenses having different focal lengths. However, in recent years, a deformed aperture stop having an aperture other than a circular shape (for example, a ring-shaped aperture or a quadrupole aperture) may be used as the aperture stop. Hereinafter, the reason why this type of modified aperture stop is used will be briefly described.

When the reticle pattern becomes fine and exposure is performed near the resolution limit of the apparatus, it is the periphery of the aperture stop that contributes to the resolution of the light flux emitted from the aperture stop of the illumination system. The light emitted from the central part of the aperture stop only has the function of lowering the contrast of the image. In other words, when transmitting reticle information to the wafer, it is only the light emitted from the peripheral portion of the aperture stop that gives the energy for transmitting information, and the light emitted from the central portion of the aperture stop generates noise, so to speak. Will only be. Therefore, it is desirable that no light be emitted from the central portion of the aperture stop.

On the basis of the above findings, a ring-shaped aperture stop having a ring-shaped opening and a 4-pole aperture stop having a 4-pole opening have been proposed. In particular, the 4-pole aperture stop
It is advantageous when the pattern to be resolved is mainly composed of vertical lines and horizontal lines. When such a modified aperture stop is used, only a part of the transmitted light amount of the fly-eye lens is used, and the light is blocked by the aperture stop in many parts including the central part near the optical axis.
This causes a loss of the amount of light in the aperture stop. As a result, the illuminance on the reticle and the wafer is lowered, which in turn lowers the throughput.

The present invention has been made in view of the above-mentioned problems, and provides a diffractive optical element and a method of manufacturing the same, which can efficiently convert an incident light beam into a light beam having a predetermined cross-sectional shape. To aim. Similarly, it is an object of the present invention to provide a mask that can be used for manufacturing by diffraction a diffractive optical element that can efficiently convert an incident light beam into a light beam having a predetermined cross-sectional shape, and a method for manufacturing the mask. It is another object of the present invention to provide an illumination optical device that includes the diffractive optical element of the present invention and is capable of forming various light intensity distributions while satisfactorily suppressing the light amount loss in the illumination pupil.

[0019]

In order to solve the above-mentioned problems, according to a first aspect of the present invention, a basic optical element composed of a plurality of partial optical elements is provided, and an incident light beam has a predetermined light beam cross section. A method of manufacturing a mask used for lithographically manufacturing an optical element for converting into a plurality of first light flux sections, the predetermined light flux section being divided into a plurality of areas extending along a first direction. And a total length calculating step of calculating a sum of lengths of the plurality of first light flux cross-section partial areas obtained in the first dividing step along the first direction, and the total length calculating step. A unit length calculation step of equally dividing the total sum of the lengths obtained by to obtain a predetermined unit length, and the plurality of first light flux cross sections based on the predetermined unit length obtained by the unit length calculation step. Divide sub-region Then, a second division step of obtaining a plurality of second light flux cross-section partial areas and coordinate data on a predetermined optical element corresponding to coordinates of the plurality of second light flux cross-section partial areas are used to determine the predetermined optical element. A third dividing step of dividing into a plurality of partial optical elements; a dense arrangement step of forming the basic optical element by densely arranging the plurality of partial optical elements divided by the third dividing step; and the dense arrangement And a drawing step of drawing on the mask a pattern for obtaining the basic optical element obtained by the step by lithography.

According to a preferred aspect of the first invention, each of the plurality of first light flux cross-section partial regions is further divided to obtain a third light flux cross-section partial region, and the subdivision process is performed. A random arrangement step of randomly rearranging a plurality of third light flux cross-section partial regions within the first light flux cross-section partial region. In addition, the plurality of third light flux cross-section partial regions have a plurality of equal division partial regions and non-equal division partial regions,
It is preferable that each of the equally divided partial areas is set to have a substantially square shape. In addition, a regular arrangement step of regularly arranging the plurality of partial optical elements divided by the third dividing step in a regular arrangement, and the plurality of partial optical elements arranged regularly and densely by the regular arrangement step are randomly arranged. It is preferable to include a random arrangement step of rearrangement.

In this case, the plurality of partial optical elements arranged by the regular arrangement step have at least one row, and the random arrangement step includes each row of the plurality of partial optical elements densely arranged. It is preferable to include a step of rearranging randomly. Further, the plurality of partial optical elements arranged by the regular arrangement step have at least two rows, and the random arrangement step randomly rearranges in a row unit in the plurality of densely arranged partial optical elements. It is preferable to include the step of further,
The plurality of partial optical elements arranged by the regular arrangement step have at least two rows, and the random arrangement step has the same size between two rows in the plurality of densely arranged partial optical elements. It is preferable to include the step of replacing the partial optical element of.

According to a preferred aspect of the first invention,
With respect to a cross section dividing step of dividing the predetermined light flux cross section into a plurality of light flux sections symmetrical to each other, and one cross section of the plurality of light flux sections symmetrical to each other obtained by the cross section dividing step, the first division step, A step of obtaining a first divided basic optical element by performing the total length calculation step, the unit length calculation step, the second division step, the third division step, and the dense arrangement step; and the first divided basic optical element Obtaining another one or a plurality of divided basic optical elements based on the above configuration and the symmetry of the plurality of light beam cross sections, and the first divided basic optical element and the other one or a plurality of divided basic optical elements. And a synthesizing step to synthesize the basic optical element.

Alternatively, according to a preferred aspect of the first invention, a portion of the partial optical element in a state where the center of the predetermined light beam cross section and the center of the predetermined optical element in the third dividing step are decentered. A shape setting step of setting a shape is included. In this case, the optical element includes a first basic optical element and a second basic optical element, and the shape setting step determines the center of the predetermined light beam cross section when obtaining the first basic optical element. A first shape setting step of setting the center of the predetermined optical element in the third division step to a first eccentric state; and a center of the predetermined light beam cross section when obtaining the second basic optical element. It is preferable to include a second shape setting step of setting the center of the predetermined optical element in the third division step to a second eccentric state.

Further, according to a preferred aspect of the first invention, the first eccentric state in the first shape setting step and the second eccentric state in the second shape setting step are symmetrical to each other. . Further, the optical element includes a first basic optical element and a second basic optical element, and the second basic optical element is a portion of the plurality of partial optical elements in the first basic optical element. It is preferably generated by performing coordinate transformation on the shape data according to a predetermined rule. further,
It is preferable that the predetermined optical element is a binary phase type Fresnel zone plate, and the basic optical element is configured based on half of the predetermined light beam cross section. In this case, the basic optical element includes a first basic optical element configured based on a first light flux cross section corresponding to a half of the predetermined light flux cross section, and the first light flux cross section rotated by 90 degrees around the optical axis. It is preferable to have a second basic optical element configured based on the second light flux cross section obtained by the above.

In the second aspect of the present invention, a basic optical element composed of a plurality of partial optical elements is provided, which is used for producing an optical element for converting an incident light beam into a light beam having a predetermined light beam cross section by lithography. A method of manufacturing a mask, comprising: a first dividing step of dividing the predetermined light beam cross section into a plurality of regions to obtain a plurality of light beam cross section partial regions; and a predetermined optical corresponding to coordinates of the plurality of light beam cross section partial regions. A second dividing step of dividing the predetermined optical element into a plurality of partial optical elements using coordinate data on the element, and the plurality of partial optical elements divided by the second dividing step are densely arranged. A dense arrangement step of forming the basic optical element,
A drawing step of drawing a pattern for obtaining the basic optical element obtained by the dense arrangement step by lithography on the mask, wherein the second dividing step includes the plurality of partial optics on the predetermined optical element. A step of determining the coordinate data of each of the elements, and a step of associating the shape data of each of the plurality of partial optical elements with the coordinate data, the dense arrangement step, the dense arrangement of the plurality of Based on the coordinate data of the partial optical element and the coordinate data in the second dividing step,
A mask manufacturing method comprising: a step of obtaining coordinate conversion data of the partial optical element; and a step of performing coordinate conversion of the shape data based on the coordinate conversion data.

According to a preferred aspect of the second invention, the second dividing step includes a step of expressing the shape data of each of the plurality of partial optical elements in a predetermined pattern.
In the step of coordinate-converting the shape data, the predetermined pattern is coordinate-converted. In this case, the second dividing step includes a mesh dividing step of dividing the plurality of partial optical elements into a large number of cells, and a plurality of cells obtained by the mesh dividing step and the plurality of partial optical elements. A determining step of determining a cell that should form a part of the pattern and a cell that does not form a part of the pattern based on the predetermined pattern, and the step of converting the coordinates of the shape data, the determining step It is preferable to coordinate-convert the cell information obtained by.

A third invention of the present invention provides a mask manufactured by the manufacturing method of the first invention or the second invention.

A fourth invention of the present invention provides an optical element manufactured by using the mask of the third invention. In this case, the optical element is preferably a diffractive optical element, a refractive optical element or a reflective optical element.

In a fifth aspect of the present invention, a basic optical element composed of a plurality of partial optical elements is used to manufacture an optical element for converting an incident light beam into a light beam having a predetermined light beam cross section by lithography. In the mask
There is provided a mask having a pattern of the basic optical element and a control regular pattern for controlling at least one of a line width and a depth.

In a sixth aspect of the present invention, there is provided a method of manufacturing an optical element comprising a basic optical element composed of a plurality of partial optical elements, for converting an incident light beam into a light beam having a predetermined light beam cross section. And a pattern transfer step of transferring the pattern on the mask to a substrate coated with a photosensitive material.

According to a preferred aspect of the sixth aspect of the invention, in the pattern transfer step, the line width of the pattern to be transferred is controlled by using the control regular pattern. Further, the method further includes a developing step of developing the substrate that has undergone the pattern transfer step, and an etching step of etching the substrate that has undergone the developing step, wherein in the etching step, the control rule pattern transferred to the substrate. It is preferable to control the etching based on the depth of. Further, the mask has a cutting guide pattern, and the pattern transfer step includes a step of exposing the cutting guide pattern to the substrate, and the pattern is transferred along the cutting guide pattern exposed on the substrate. It is preferable to further include a cutting step of cutting the substrate.

The seventh invention of the present invention provides an optical element manufactured by the manufacturing method of the sixth invention.

According to an eighth aspect of the present invention, a basic optical element composed of a plurality of partial optical elements is provided, the plurality of partial optical elements convert an incident light beam into a plurality of partial light beams, and the basic optical element is incident. An optical element for converting a light flux into a light flux having a predetermined light flux cross section as a sum of the plurality of partial light fluxes, wherein the basic optical element corresponds to a plurality of partial regions obtained by dividing the predetermined light flux cross section. Provided is an optical element having a characteristic that the plurality of partial optical elements set by the above are arranged densely and randomly.

According to a preferred aspect of the fourth invention, the seventh invention or the eighth invention, the plurality of partial optical elements have a diffraction pattern, and the line width of the diffraction pattern is more than 0.1 μm and more than 1000 μm. Is also small. More preferably,
It is larger than 0.1 μm and smaller than 200 μm.

According to a ninth aspect of the present invention, in an illuminating optical device for illuminating a mask on which a predetermined pattern is formed, a light source for supplying an illuminating light beam and an optical element according to the fourth, seventh or eighth aspect of the invention. Based on the light flux from the light flux conversion section including an element, the light flux conversion section for converting the illumination light flux from the light source into a light flux having a predetermined light flux cross section and guiding the light flux to a predetermined surface, and the light flux conversion section having the predetermined light flux cross section. There is provided an illumination optical device comprising: an optical integrator for forming a surface light source having a predetermined shape; and a light guide optical system for guiding a light beam from the optical integrator to the mask.

According to a preferred aspect of the ninth aspect of the invention, the optical element in the light flux converting section includes a plurality of the basic optical elements. In this case, the plurality of basic optical elements,
A first basic optical element and a second basic optical element, wherein the first basic optical element and the second basic optical element are each a plurality of partial regions obtained by dividing the predetermined light beam cross section. The plurality of partial optical elements set corresponding to, are densely and randomly arranged, and are configured,
The random arrangement of the plurality of partial optical elements in the first basic optical element and the random arrangement of the plurality of partial optical elements in the second basic optical element are preferably different from each other. Further, it is preferable that the light flux conversion section includes a plurality of optical elements that can be switched with respect to the illumination optical path in order to change the light intensity distribution on the predetermined surface. Further, the optical integrator has a wavefront division type optical integrator having a plurality of optical elements arranged two-dimensionally, the division pitch for dividing the predetermined light beam cross section into the plurality of first light beam cross section partial regions is , Is preferably set based on the arrangement pitch of the optical elements. Further, in order to set at least one of the illuminance distribution on the conjugate plane of the mask and the illuminance distribution on the illumination pupil plane to a desired state, the basic optical element is an array of the plurality of densely arranged partial optical elements. Is set as an initial value and is set by optimizing the arrangement of the partial optical elements using a predetermined algorithm. Furthermore, in order to set at least one of the illuminance distribution on the conjugate plane of the mask and the illuminance distribution on the illumination pupil plane to a desired state, the basic optical element is an array of the plurality of densely arranged partial optical elements. And an internal pattern as an initial value, each partial optical element is further subdivided, and the setting is performed by optimizing the arrangement of the partial optical elements and the arrangement of the subdivided internal patterns using a predetermined algorithm. preferable.

A tenth aspect of the present invention is characterized by comprising the illumination optical device of the ninth aspect and a projection optical system for projecting the pattern image of the mask onto a substrate coated with a photosensitive material. An exposure apparatus is provided.

An eleventh aspect of the present invention includes an illumination step of illuminating the mask with the illumination optical device of the ninth aspect, and a projection step of projecting a pattern image of the mask onto a substrate coated with a photosensitive material. An exposure method characterized by the above.

According to a twelfth aspect of the present invention, in a mask in which a basic pattern to be exposed is drawn on a substrate coated with a photosensitive material, a regular pattern for controlling the line width of the pattern exposed on the substrate is provided. And a cutting guide pattern for guiding the cutting of the substrate is drawn in parallel.

According to a preferred aspect of the twelfth invention, the regular pattern is a line-and-space pattern or a dot pattern. Further, it is preferable that the mask is used to etch the substrate on which the basic pattern is exposed to form a desired pattern, and the regular pattern is used to control the etching depth.

[0041]

BEST MODE FOR CARRYING OUT THE INVENTION A diffractive optical element as a typical optical element according to the present invention comprises a basic optical element composed of a plurality of partial optical elements. Here, the plurality of partial optical elements convert the incident light flux into a plurality of partial light fluxes, and the basic optical element converts the incident light flux into a light flux having a predetermined light flux cross section as a sum of the plurality of partial light fluxes. The basic optical element is configured by densely arranging a plurality of partial optical elements set corresponding to a plurality of partial regions obtained by dividing a predetermined light beam cross section. When a plurality of basic optical elements are provided, each basic optical element converts the incident light flux into a light flux having a predetermined light flux cross section and superimposes them.

FIG. 1 is a flow chart showing a manufacturing flow of a mask used for manufacturing an optical element according to the present invention by lithography. Hereinafter, a method for manufacturing a mask according to the present invention, and further a method for manufacturing an optical element according to the present invention, will be described with reference to FIG. In the present invention, a predetermined light beam cross section is divided into a plurality of regions extending along the first direction to obtain a plurality of first light beam cross section partial regions (S1).
1: the first dividing step). FIG. 2 is a diagram for explaining the first dividing step in the present invention.

In FIG. 2, each of the regions R11 to R1n indicated by the rectangular hatched portion corresponds to n (12 in FIG. 2) first light beam cross-sectional partial regions. A ring-shaped region R0 (region indicated by hatching in the drawing) indicated by the entire n first light flux cross-section partial regions R11 to R1n corresponds to a predetermined light flux cross section (that is, a desired light flux cross section). In other words, in FIG. 2, the n-shaped first luminous flux cross-section R0 to be formed via the optical element is divided into the n first luminous flux sections R0 extending in the lateral direction in the figure.
It is divided into light flux cross-section partial regions R11 to R1n. The dimensions of the first light flux cross-section partial regions R11 to R1n along the vertical direction in the drawing are set to be equal to each other.

Next, in the present invention, the n first light flux cross-section partial regions R11 to R1 obtained in the first dividing step S11 are obtained.
The sum of the lengths of n along the first direction (horizontal direction in FIG. 2) is calculated (S12: total length calculation step). FIG. 3 is a diagram for explaining the total length calculation step in the present invention. As shown in FIG. 3, each of the first light flux cross-section partial regions R11 to R1n has a length X1 to Xn along the lateral direction in the drawing. Therefore, the sum Lx_tot of the lengths X1 to Xn of the n first light flux cross-section partial regions R11 to R1n is Lx_tot = Σ.
(I = 1 to n) is given by Xi. Where Σ (i = 1
~ N) are the summation symbols for i = 1 to n.

Next, in the present invention, the total length Lx_tot obtained in the total length calculation step S12 is equally divided to obtain a predetermined unit length (S13: unit length calculation step). In the unit length calculation step S13, the number of divisions my (integer) into which the total Lx_tot should be equally divided is set. The meaning of the division number my will be described later. Thus, the unit length Lx_rea is
Based on the sum Lx_tot and the number of divisions my, Lx_r
It is given by ea = Lx_tot / my.

Next, in the present invention, the unit length calculation step S13
Based on the unit length Lx_rea obtained by, the n first light flux cross-section partial regions R11 to R1n are divided to obtain m second light flux cross-section partial regions R21 to R2m (S14: second division step). FIG. 4 is a diagram for explaining the second dividing step in the present invention. In FIG. 4, for example, the first light flux cross-section partial region R11 includes second light flux cross-section partial regions R21 and R22 having a unit length Lx_rea, and a unit length Lx_re.
The second light beam cross-section partial region R23, which is a surplus element in the division into a. The other first light flux cross-section partial regions R12 to R1n are similarly divided based on the unit length Lx_rea.

Next, in the present invention, by using coordinate data on a predetermined optical element corresponding to the coordinates of the m second light flux cross-sectional partial regions R21 to R2m, the predetermined optical element is converted into m partial optics. Divide into elements (S15: third dividing step). Figure 5
FIG. 6 is a diagram illustrating a third division step in the present invention.
In FIG. 5, the predetermined optical element OP is divided into m partial optical elements OP1 to OPm. Here, the second light beam cross-section partial region R2i (i = 1 to m) and the partial optical element OPi
(I = 1 to m) optically corresponds via the optical system LS.

Next, in the present invention, the m optical elements OP1 to OPm divided in the third dividing step S15 are densely rearranged to form a basic optical element (S16: dense arrangement step). FIG. 6 is a diagram for explaining the dense arrangement step in the present invention. In FIG. 6, m partial optical elements OP1
~ OPm are densely arranged to form a rectangular basic optical element BO
Are configured.

Next, in the present invention, a pattern for obtaining the basic optical element BO obtained by the dense arrangement step S16 by lithography is drawn on a mask (S17: drawing step). Specifically, with a well-known reticle (mask) drawing device, the pattern corresponding to the basic optical element BO is
EB drawing or laser drawing is performed on the mask. Thus
The mask used to lithographically manufacture the optical element of the present invention is manufactured.

Next, in the present invention, the pattern on the mask obtained in the drawing step S17 is transferred to the substrate coated with the photosensitive material (S18: pattern transfer step). The pattern transfer step S18 is repeated by the number of basic optical elements BO that form an optical element, for example. Further, the substrate that has undergone the pattern transfer step S18 is developed (S19: developing step). Finally, the substrate that has undergone the developing step S19 is etched (S110: etching step). Thus, the optical element of the present invention is manufactured.

As described above, in the optical element according to the present invention, the plurality of densely arranged partial optical elements convert the incident light beam into a plurality of partial light beams. Then, the basic optical element, which is the sum of the plurality of densely arranged partial optical elements,
The incident light flux is converted into a light flux having a predetermined light flux cross section as a sum of a plurality of partial light fluxes. As a result, with the optical element of the present invention, it is possible to efficiently convert the incident light beam into a light beam having a predetermined cross-sectional shape (while suppressing the light amount loss satisfactorily).
Further, in the illumination optical device including the optical element of the present invention, by switching a plurality of optical elements, various light intensity distributions can be formed while satisfactorily suppressing the light amount loss in the illumination pupil.

Embodiments of the present invention will be described with reference to the accompanying drawings. FIG. 7 is a diagram schematically showing a configuration of an exposure apparatus equipped with an illumination optical device including a diffractive optical element according to each embodiment of the present invention. Further, FIG. 8 shows the first part of FIG.
It is a figure which shows the some diffractive optical element provided in the revolver, and the some aperture stop provided in the 2nd revolver. In FIG. 7, the z axis is along the optical axis of the projection optical system, the y axis is parallel to the plane of FIG. 7 in the plane perpendicular to the z axis, and the x axis is perpendicular to the plane of FIG. 7 in the plane perpendicular to the z axis. Each axis is set.

In the exposure apparatus shown in FIG. 7, after the luminous flux from the excimer laser light source 1 is changed by the reflection mirror M1,
It is incident on the beam shaping optical system 2. The light flux converted into a desired cross-sectional shape via the beam shaping optical system 2 is revolver 3
The light is incident on the diffractive optical element 21 provided in, and a desired secondary light source array is formed near the exit surface thereof. The light flux from the secondary light source array is superimposed on the incident surface of the fly-eye integrator (fly-eye lens) 6 with a desired light intensity distribution via the relay lens system 4 and the oscillating mirror 5 arranged in the optical path thereof. Uniform illumination. As a result, a substantial surface light source is formed as a tertiary light source on or near the exit surface of each lens element of the fly-eye integrator 6.

The light flux from the surface light source formed on the exit surface of each lens element of the fly-eye integrator 6 or in the vicinity thereof is limited in shape by the aperture stop 33 provided in the revolver 7, and then the condenser. The reticle (mask) 9 on which the pattern is drawn is superimposed and uniformly illuminated via the lens system 8 and the reflection mirror M2 arranged in the optical path thereof. The pattern formed on the reticle 9 is projected and exposed through the projection optical system 10 onto each exposure area on the substrate, that is, the wafer 11 coated with the photosensitive material.

The exposure to each exposure area is performed by combining exposures of several tens of pulses. However, by changing the angle of the vibrating mirror 5 over the exposure time, the fly-eye integrator 6 or the like causes the exposure. The generated interference noise is averaged. Further, the combined optical system of the relay lens system 4 and the fly-eye integrator 6 constitutes an imaging optical system,
The entire effective diameter in the xy plane of the secondary light source array near the exit surface of the diffractive optical element 21 is set so as to form an image as a tertiary light source array near the exit surface of each lens element of the fly-eye integrator 6. There is.

Further, as shown in FIG. 8A, the revolver 3 is provided with six diffractive optical elements 21 to 26 and two auxiliary fly-eye integrators 27 and 28. Note that the diffractive optical elements 25 and 26 are shown separately from the revolver 3 for the sake of clarity of the drawing. Then, by rotating the revolver 3 about the rotation axis parallel to the optical axis AX by the action of the motor MT1,
The desired diffractive optical element or auxiliary fly's eye integrator is selectively set in the illumination beam path.

Further, the revolver 7 is provided with eight aperture stops 31 to 38 as shown in FIG. 8 (b).
Note that the aperture stops 37 and 38 are shown for clarity of the drawing.
Is shown separately from the revolver 7. And the motor MT
The desired aperture stop is selectively set in the illumination optical path by rotating the revolver 7 about the rotation axis parallel to the optical axis AX by the action of 2. The aperture stop 3
Reference numerals 2, 34 and 35 are circular aperture stops having a circular opening, and aperture stops 33 and 36 are annular aperture stops having an annular opening.

The aperture stop 31 has a quadrupole shape (fourth shape).
Is a four-pole aperture stop, and the aperture stops 37 and 38 are two-pole (second-shaped) aperture stops. When the auxiliary fly-eye integrator 27 or 28 is set in the illumination optical path by the rotation of the revolver 3, it becomes a conventional double fly-eye lens system, and the auxiliary fly-eye A large number of tertiary light sources corresponding to the product m × n of the number m of the lens elements of the integrator and the number n of the lens elements of the fly-eye integrator 6 are formed.

Here, according to the setting of the auxiliary fly-eye integrator 27 in the illumination optical path, a circular aperture stop 32 having a medium aperture diameter is set in the illumination optical path. Further, in accordance with the setting of the auxiliary fly-eye integrator 28 in the illumination light path, the circular aperture stop 34 having a relatively small aperture diameter is set in the illumination light path. Further, according to the setting of the diffractive optical element 22 for circular illumination in the illumination optical path, a circular aperture stop 35 having a relatively large aperture diameter is set in the illumination optical path. In addition, the diffractive optical element 21 for annular illumination
In accordance with the setting to the illumination optical path, an annular aperture stop 33 having an annular opening having a relatively large outer diameter is set in the illumination optical path.

Further, according to the setting of the diffractive optical element 23 for annular illumination in the illumination optical path, an annular aperture stop 36 having an annular opening having a relatively small outer diameter is set in the illumination optical path. . Further, in accordance with the setting of the diffractive optical element 24 for quadrupole illumination in the illumination optical path, four circular openings 4 are provided.
A polar aperture stop 31 is set in the illumination optical path. Furthermore, 2
In accordance with the setting of the diffractive optical element 25 for polar illumination in the illumination optical path, a two-pole aperture stop 37 having two circular openings arranged side by side in the drawing is set in the illumination optical path. Also, 2
In accordance with the setting of the diffractive optical element 26 for polar illumination in the illumination optical path, a two-pole aperture stop 38 having two circular openings arranged in the vertical direction in the figure is set in the illumination optical path.

In this way, scan exposure is performed while two-dimensionally drivingly controlling the wafer 11 in a plane (xy plane) orthogonal to the optical axis AX of the projection optical system 10.
The pattern of the reticle 9 is sequentially exposed in each exposure area of the wafer 11. In scan exposure, the so-called step
According to the AND-scan method, the reticle pattern is scan-exposed on each exposure area of the wafer 11 while moving the reticle 9 and the wafer 11 relative to the projection optical system 10 along the y direction (scan direction). In this case, the shape of the illumination area on the reticle 9 is a rectangular shape in which the ratio of the short side to the long side is, for example, 1: 3.
The cross-sectional shape of each lens element of the integrator 6 is also a rectangular shape similar to this.

(First Embodiment) As a first embodiment, the structure of a diffractive optical element 21 for annular illumination used together with an aperture stop 33 for annular illumination and a method of manufacturing the same will be described below. FIG. 9 is a diagram showing the relationship between the fly-eye integrator and the ring-shaped aperture of the ring aperture stop in the first embodiment. In order to facilitate understanding of the present invention, in the drawings after FIG. 9, it is assumed that the optical path in FIG. 7 extends straight without being bent by the vibrating mirror 5 and the reflecting mirror M2, and the coordinates x and y. Is shown.

As shown in FIG. 9, the fly-eye integrator 6 is constructed by arranging a large number of lens elements EL having a long and narrow rectangular cross section along the x direction (non-scanning direction) vertically and horizontally and densely. ing. Then, in the two columns adjacent to each other in the y direction (scanning direction), the pitch of the lens elements EL along the x direction is displaced by 1/2 pitch. The center of the fly-eye integrator 6 and the center of the ring-shaped aperture of the ring-shaped aperture stop 33, that is, the center of the transmission area AP1 coincide with the optical axis AX.

FIG. 10 is a diagram showing the relationship between an annular illumination region to be formed on the incident surface of the fly-eye integrator and a large number of lens elements in the first embodiment. Referring to FIG. 10, a part of a large number of lens elements EL constituting the fly-eye integrator 6 relevant to the description, an annular transmission area AP1 of the aperture stop 33 (indicated by a broken line in the figure), and , A ring-shaped basic illumination area (indicated by a broken line in the figure) IA1 to be formed on the incident surface of the fly-eye integrator 6. The rectangular area SP drawn inside the rectangular area of each lens element EL is an area obtained by proportionally reducing the rectangular area of each lens element EL, and the emission of the fly-eye integrator 6 is performed. It is a light spot region formed on the side.

In FIG. 10, it is assumed that each light spot area is reduced to 1/2 with respect to the rectangular area of each lens element EL, that is, the filling rate of the light spot area SP is 50%. ing. Here, the lens element EL whose light spot area SP is included in the transmission area AP1 of the aperture stop 33 is a necessary lens element, and FIG. 10 illustrates part of the necessary lens element. On the other hand, the basic illumination area IA
1 is set by adding a margin portion to the outermost periphery of the assembly of necessary lens elements. In addition, in FIG. 10, when setting the basic illumination area IA1, the lens element having a very small transmission contribution (for example, indicated by reference numeral EL9) is removed from the necessary lens elements.

However, when the output power of the light source 1 has a margin and priority is given to the uniformity of illumination, it is possible to set the basic illumination area IA1 with all the lens elements having a very small transmission contribution included. Is. Further, since the outer diameter tolerance of the transmissive area AP1 of the aperture stop 33, the setting position accuracy of the transmissive area AP1 and the alignment accuracy of the basic illumination area IA1 and the fly-eye integrator 6 are not zero, they are necessary for the tolerance. It is also possible to set the lens element or the basic illumination area IA1 to be wider. It is desirable that the setting of the necessary lens elements and the setting of the basic illumination area IA1 based on the necessary lens elements are finally set by simulation or the like so that the illumination uniformity is satisfied after including various tolerances.

FIG. 11 is a diagram showing a projected basic area obtained by projecting the basic illumination area onto a predetermined virtual diffraction lens. FIG. 12 is a diagram for explaining the relationship between the virtual diffractive lens virtually arranged at the position of the diffractive optical element and the basic optical element. First, referring to FIG. 12, the virtual diffractive lens 47 is assumed on the surface of the diffractive optical element 21. Then, the focal length of the virtual diffraction lens 47 is set to funi.
And the focal length of the relay lens system 4 is fL,
The virtual diffraction lens 47 and the relay lens system 4 are (funi
+ FL), and the combined system of the virtual diffraction lens 47 and the relay lens system 4 constitutes an image forming optical system.
Note that the relay lens system 4 is schematically represented by a single lens for clarity of the drawing.

Further, in FIG. 12, a projection basic region 48 is a region obtained by projecting the basic illumination region 45 on the incident surface of the fly-eye integrator 6 onto the image forming surface on the side of the virtual diffractive lens via this imaging optical system. I am trying. That is, since the virtual diffractive lens 47 illuminates the basic illumination area 45 on the incident surface of the fly-eye integrator 6, it has a function equivalent to that of the basic optical element. Further, similarly to the basic illumination area, the partial illumination area 46 also corresponds to the projection partial area 49. Here, since the light beam incident on the virtual diffractive lens 47 is substantially parallel light, each light beam corresponding to the projection partial region 49 goes straight on and enters the virtual diffractive lens 47.

Then, on the virtual diffraction lens 47,
A region of the projection partial region 49 on which each light beam is incident is considered as a partial optical element 50. As a result, the pattern of the partial illumination area 46 on the incident surface of the fly-eye integrator 6 is projected on the virtual diffractive lens 47 with a predetermined similarity ratio, and the projected partial illumination area pattern of the virtual diffractive lens 47 is projected. Corresponds to the partial optical element. However, the partial optical element required for annular illumination is a portion corresponding to the partial illumination area 46,
The partial optical elements in the central portion and the peripheral portion are unnecessary portions.
Therefore, as will be described later, the partial optical elements unnecessary for annular illumination are removed and only the necessary partial optical elements are selected,
Further, the partial optical elements selected so that they can be illuminated efficiently can be re-arranged in a dense state to form a new basic optical element.

Next, referring to FIG. 11, a projection basic area PIA obtained by projecting the basic illumination area IA1 onto a virtual diffractive lens virtually arranged at the position of the diffractive optical element 21.
Only the upper half divided projection basic region of 0 ≦ y of 1 is shown. As described above, since the virtual diffractive lens portion included in the projection basic area PIA1 corresponds to the basic optical element, the projection basic area PIA1 and the basic optical element can be identified. Further, as the virtual diffractive lens, for example, a phase type binary Fresnel zone plate (or a multi-stage type) can be used.

Since the partial optical element is obtained by dividing the virtual diffractive lens, the virtual diffractive lens constitutes a "predetermined optical element" which is the basis for obtaining each partial optical element by division. There is. In general, "given optics"
As the element, a diffractive optical element, a refractive optical element, or a reflective optical element can be used. When a diffractive optical element is used as the "predetermined optical element", for example, a diffraction grating can be used instead of the virtual diffractive lens. For more information on diffractive optical elements as "predetermined optical elements",
Reference can be made to the specification and the drawings of Japanese Patent Application No. 2000-35910.

When the phase type binary Fresnel zone plate is used, the radius d _{m of} the virtual diffractive lens with respect to its origin (center point) is expressed by the following equation (1), and r _m and r
The phase of the annular zone surrounded by _{m + 1} and is 0, and r
_The diffractive lens is such that the phase of the zone surrounded by _{m + 1} and r _{m + 2} is λ / 2. Where λ is the wavelength of light,
f is the focal length of the virtual diffractive lens, and m is an arbitrary integer. _{= r m [(m · λ} ) 2/4 + m · f · λ] 1/2 (1)

Such a phase diffractive lens has r _m and r _{m + 1}
A zone plate original plate (mask) was prepared in which a zone surrounded by and a light-shielding portion and a zone surrounded by r _{m + 1} and r _{m + 2} became a transparent portion, and the mask pattern was coated with a resist. It can be formed by exposing a glass substrate, developing it, and etching it. Hereinafter, for simplicity of description, the above-described pattern is simply referred to as an FZP pattern (Fresnel zone plate pattern).

FIG. 13A is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions.
3B is a diagram showing a divided basic optical element obtained by densely rearranging a large number of partial optical elements defined corresponding to a large number of divided partial areas. In addition, FIG.
It is a figure explaining the division rule in 3 (a). FIG.
In (a), the upper half divided projection basic area PIA1 is x
It is divided into 51 partial regions extending along the direction. Here, each partial region has the same dimension q along the y direction.
have y. That is, in the first embodiment, the divided projection basic area PIA1 is divided along the y direction at equal intervals qy (see FIG. 14).

Next, referring to FIG. 14, partial regions divided at equal intervals qy along the y direction are arranged in a line along the x direction, and the total length in the x direction is Lx_tot. Then, a value obtained by dividing the total length Lx_tot by the integer my is set as a unit length Lx_rea (Lx_rea = Lx_tot /
my). In the first embodiment, the division number my is set to 16 as an example.
This corresponds to the number of steps of the lens element in the y direction in the corresponding region (upper half region) of the integrator 6.
Next, a new dividing line 41 is inserted at a position where the total length Lx_tot is divided by the unit length Lx_rea.

Thus, as shown in FIG. 14, in addition to the original dividing line, that is, the dividing line 40 in the first dividing step,
A new dividing line, that is, a dividing line 41 by the second dividing process is formed. By the above procedure, 51 partial areas shown in FIG. 13A are set. And FIG. 13 (a)
The 51 partial optical elements can be obtained by overlapping the 51 partial areas shown in 1 above with the FZP pattern described above, in other words, by dividing the FZP pattern according to the 51 partial areas. In FIG. 13B, 51 partial optical elements defined corresponding to the 51 partial regions are densely rearranged, the dimension along the x direction is Lx_rea, and the dimension along the y direction is ( 16 x qy)
The rectangular divided basic optical element A + is formed.

The projection basic area PIA1 shown in FIG.
13B to the dense arrangement shown in FIG. 13B, the coordinate system O of FIG. 13A and the coordinate system OR of FIG. The position of each defined partial optical element (including FZP pattern information) can be converted into the position of each defined partial optical element defined in the coordinate system OR and arranged densely. For example, all coordinates of the data points included in the partial optical element 1 of FIG. 13A are coordinate-converted to the position of the partial optical element 1 of FIG. 13B. Then, this operation may be performed for all the elements from the partial optical elements 1 to 51. As basic data therefor, the coordinate conversion relationship between the coordinate system O and the coordinate system OR of the four vertexes of the rectangular regions of the partial optical elements 1 to 51 is tabulated, and based on the basic number table, You may use the procedure of coordinate-converting a data point.

FIG. 15A is a diagram showing a state in which the lower half divided projection basic region is divided into a large number of partial regions.
FIG. 5B is a diagram showing a divided basic optical element obtained by densely rearranging a large number of partial optical elements defined corresponding to a large number of divided partial areas. As shown in FIG. 15 (a), the lower half divided projection basic region of y <0 can be similarly divided into 51 partial regions, and 51 partial regions corresponding to the 51 partial regions can be similarly divided. An optical element is obtained. Then, as shown in FIG. 15B, 51 partial optical elements defined corresponding to the 51 partial regions are densely rearranged, and the dimension along the x direction is Lx_rea.
Thus, a rectangular divided basic optical element A- having a dimension (16 × qy) along the y direction is formed.

FIG. 16 is a diagram showing how the basic optical element is formed by synthesizing the basic divided optical element shown in FIG. 13B and the basic divided optical element shown in FIG. 15B. As shown in FIG. 16, in the incident surface of the fly-eye integrator 6, a ring-shaped basic illumination area IA1 as shown in FIG.
The basic optical element A for forming the whole of FIG.
It is obtained by arranging the divided basic optical element A + of (b) and the divided basic optical element A- of FIG. 15 (b) adjacently in the y direction. However, one set of the divided basic optical element A + and the divided basic optical element A- is not limited to this, and various modified examples are possible.

That is, the divided basic optical element A + and the divided basic optical element A- can be arranged adjacent to each other according to various forms, or the divided basic optical element A + and the divided basic optical element A- can be arranged in various forms. According to the above, they can be spaced apart. As a coordinate conversion method for one set, the divided basic optical element A + and the divided basic optical element A- may be designed by different coordinate systems OR, and then the coordinate conversion may be performed again so that they are adjacent to each other. However, a common coordinate system OR may be set such that the divided basic optical element A + and the divided basic optical element A− are originally adjacent to each other.

In the first embodiment, a pattern for obtaining the designed basic optical element A by lithography is drawn on a mask. Specifically, using a well-known reticle (mask) drawing device, the pattern corresponding to the basic optical element A is EB drawn or laser drawn on the mask. In this way, the mask used for manufacturing the diffractive optical element 21 of the first embodiment by lithography is created. Next, the created pattern on the mask is transferred to the substrate coated with the photosensitive material by, for example, reduction projection exposure. This pattern transfer is repeated by the number of basic optical elements forming the diffractive optical element 21. further,
The substrate that has undergone pattern transfer is developed, and the developed substrate is etched. In this way, the diffractive optical element 21 of the first embodiment is manufactured. In the diffractive optical element, the line width of the diffraction pattern formed on each partial optical element is 0.1 μm.
It is desirable that it is larger than m and smaller than 1000 μm. Further, it is more desirable that the line width of the diffraction pattern formed on each partial optical element in the diffractive optical element is larger than 0.1 μm and smaller than 200 μm.

When a mask as a pattern master of a diffractive optical element is subjected to reduction projection exposure on a glass substrate, the size of the projection basic area PIA1 is increased by the projection magnification mag when designing the basic optical element pattern to be formed on the mask. Just set it. Similarly, the zone radius r ′ _m, which is proportionally enlarged by the projection magnification mag, with respect to the zone radius r _m of the actual virtual diffractive lens may be set as shown in the following expression (2). _{r 'm = mag · r m} = mag · [(mλ) 2/4 + mfλ] 1/2 (2)

That is, when a diffractive optical element is manufactured by subjecting a mask pattern to reduction projection exposure on a substrate, a coordinate space proportionally enlarged by the projection magnification mag is set to design the above-mentioned basic optical element. The method can be directly applied to the design of the basic optical element pattern of the mask. In addition, when using equal-magnification exposure such as proximity exposure, the projection magnification mag = 1,
The design of the basic optical element pattern of the mask results in the design of the basic optical element.

By the way, in the first embodiment, the position of the basic illumination area is slightly different to suppress uneven illumination on the entrance surface of the fly-eye integrator 6 satisfactorily.
The diffractive optical element 21 is designed to include a basic optical element of a type. Hereinafter, it will be briefly described that the unevenness of the illumination on the entrance surface of the fly-eye integrator 6 can be favorably suppressed by slightly changing the positions of the basic illumination regions.

For simplicity, the basic optical element A and the basic optical element A are arranged so that the illumination area formed by the basic optical element A and the illumination area formed by the basic optical element B are slightly displaced on the incident surface of the fly-eye integrator 6. Design B. In this case, the uniformity of the composite intensity distribution of the basic optical elements A and B is improved by setting the positional deviation amount of the illumination area so that the peaks and valleys of the noise pattern in the illumination area fill each other. You can When the noise pattern is generated two-dimensionally, it is desirable that the four illumination areas formed by four types of basic optical elements are two-dimensionally displaced.

Here, the intensity distribution of the illumination area formed by the basic optical element A is obtained by the wave-optical simulation, and the intensity pattern on the incident surface of the fly-eye integrator 6 is obtained, whereby the position shift of the illumination area is obtained. The amount can be calculated. It should be noted that when a prototype is actually manufactured, noise patterns due to manufacturing errors may overlap, so the intensity distribution is measured, and the position of the illumination area formed by the basic optical element is adjusted so as to cancel the desired noise pattern. It is desirable to redesign. For details of the technique for suppressing illumination unevenness satisfactorily due to displacement of the illumination area formed by each basic optical element, see Japanese Patent Application No. 2000
-35910 Specification and drawings may be referred to.

In the first embodiment, the center of the projection basic area and the center of the FZP pattern are gradually decentered,
The partial optical element is defined by dividing the FZP pattern. FIG. 17 is a diagram showing a state in which the center of the projection basic region and the center of the FZP pattern are eccentric.
Referring to FIG. 17, the center O of the projection illumination area PIA1 and the center O that defines the ring-shaped zone radius of the FZP pattern.
z (center of the zone plate) is set to be eccentric by Δx and Δy along the x and y directions, respectively.

Due to this eccentricity setting, the projection illumination area PIA1 and the FZP pattern have a shifted positional relationship with each other, as shown in FIG. In this case, since the FZP pattern and the projection illumination area PIA1 are eccentric, the basic optical element composed of a large number of partial optical elements obtained by dividing the projection illumination area is on the incident surface of the fly-eye integrator 6. The center of the basic illumination area formed in 1 is also decentered from the optical axis AX. Then, the eccentricity amount and the eccentric direction of the basic illumination area on the incident surface of the fly-eye integrator 6 are the projection illumination area PIA of the center Oz of the FZP pattern.
It can be set by the eccentricity amounts Δx and Δy with respect to the center O of 1.

Here, the necessary decentering amounts of the basic illumination area on the incident surface of the fly-eye integrator 6 are Δx_f and Δy_f, and the converted necessary decentering amounts on the actual diffractive optical element are Δx_d and Δy_d, Letting Δx and Δy be the converted necessary decentering amounts on the mask for manufacturing the diffractive optical element by reduction projection exposure, the following formulas (3) and (4) are provided between these decentering amounts. There is a relationship shown (shown by a two-dimensional vector display). (Δx_d, Δy_d) = funi / fL · (Δx_f, Δy_f) (3) (Δx, Δy) = mag · (Δx_d, Δy_d) (4)

Here, funi is the focal length of the virtual diffraction lens, fL is the focal length of the relay lens system 4, and mag is the projection magnification in reduction projection exposure. As shown in Expression (4), when a diffractive optical element is manufactured by reduction projection exposure using a mask, the eccentricity amounts Δx and Δy on the mask must be corrected by the projection magnification mag. Should be noted. The required eccentricity amounts Δx_f and Δy_f on the incident surface of the fly-eye integrator 6 can be obtained from lighting simulations, experimental results, and the like. Thus, the converted required eccentricity amounts Δx and Δy on the mask are obtained by using the above equations (3) and (4).

FIG. 18 shows an arrangement of four types of basic optical element patterns formed on a mask used for manufacturing a diffractive optical element including four types of basic optical elements in which the positions of basic illumination areas are slightly different from each other. FIG.
Referring to FIG. 18, four types of basic optical element patterns A to D are arranged on the mask so as to be adjacent to each other. Here, the decentering amounts Δx and Δy in the basic optical element patterns A to D are given by the following equations (5) to (8), respectively, with a and b being the optimum values obtained by the simulation.

[0092] A: (Δx, Δy) = (a, b) (5) B: (Δx, Δy) = (− a, b) (6) C: (Δx, Δy) = (a, −b) (7) D: (Δx, Δy) = (− a, −b) (8)

As a method of designing the basic optical element patterns A to D on the mask when the eccentricity as shown in the equations (5) to (8) is present, the zone plate for each basic optical element pattern A to D is used. Although the eccentricity can be set individually, the basic optical element patterns A to B can be designed individually by applying the vertical inversion rule, the horizontal inversion rule, and the vertical and horizontal inversion rule to the basic optical element pattern A. You can also ask. FIG. 19 is a diagram for explaining application of the vertical inversion rule, the horizontal inversion rule, and the vertical and horizontal inversion rule to the basic optical element pattern.

Referring to FIG. 19, for example, (Δx, Δ
y) = (a, b) The basic optical element pattern A_x generated by subjecting the basic optical element pattern A designed on the basis of the eccentricity of the zone plate to the inversion in the x direction (i) , Basic optical element pattern B
Matches In addition, in the basic optical element pattern A, (2)
The basic optical element pattern A_y generated by performing inversion in the y direction, that is, upside down, coincides with the basic optical element pattern C. Furthermore, as shown in (3),
In the basic optical element pattern A_x generated by performing the left-right reversal of (1) on the basic optical element pattern A,
The basic optical element pattern A_xy generated by performing the vertical inversion of (2) matches the basic optical element pattern D.

When the method shown in FIG. 19 is used, it is only the basic optical element pattern A that is actually designed through the division and the dense arrangement, and the other basic optical element patterns B to
There is an advantage that the data of D can be determined according to a simple rule called the inversion rule. In FIG. 18, basic optical element patterns A and C and B and D
And are shifted by sy along the y direction, which is a method for averaging interference noise in the scanning direction.

When the fly-eye integrator 6 is illuminated with a coherent light beam, the interference noise occurs when the light beam is wavefront-divided and recombined at the wafer conjugate plane. In this case, it is possible to reduce illumination nonuniformity caused by interference noise by devising a method of arranging the plurality of basic optical elements. Specifically, in the case of the first embodiment, the interference noises in the scanning direction can be averaged by shifting the basic optical element patterns A and C and B and D by sy along the y direction. . For details of averaging interference noise in the scan direction, see Japanese Patent Application No. 2000-359.
Reference may be made to the ten specifications and drawings.

In the above description, an example including four types of basic optical elements in which the positions of the basic illumination areas are slightly different, that is, an example in which the number of overlaps is four, is shown. It is also possible to set the number of laps larger than four. 20A shows the arrangement of nine types of basic optical elements when the number of overlaps is 9, and FIG.
It is a figure which respectively shows arrangement | positioning of 16 types of basic optical elements in the case of.

In FIG. 20A, what is actually designed through the division and the dense arrangement is the basic optical element patterns A to
Only D. Then, the basic optical element pattern A_x
Is generated by subjecting the basic optical element pattern A to x-direction front-back inversion, that is, left-right inversion. The basic optical element pattern A_y is generated by subjecting the basic optical element pattern A to face-to-face inversion in the y direction, that is, upside down. The basic optical element pattern B_y is the basic optical element pattern B
It is generated by upside down. The basic optical element pattern C_x is generated by performing left-right inversion on the basic optical element pattern C. Basic optical element pattern A
_Xy is generated by vertically inverting the basic optical element pattern A_x.

Also in FIG. 20B, only the basic optical element patterns A to D are actually designed through the division and the dense arrangement. Then, the basic optical element pattern A
_X to D_x are respectively generated by subjecting the basic optical element patterns A to D to front-back inversion in the x direction, that is, left-right inversion. The basic optical element patterns A_y to D_y are respectively generated by subjecting the basic optical element patterns A to D to the y-direction front-back inversion, that is, the vertical inversion. The basic optical element patterns A_xy to D_xy are respectively generated by vertically inverting the basic optical element patterns A_x to D_x.

FIG. 21 is a diagram schematically showing a structure of a mask used for manufacturing the diffractive optical element in the first embodiment by lithography. In addition, FIG.
FIG. 22 is a diagram showing one diffractive optical element among a plurality of diffractive optical elements generated on a glass substrate using the mask of FIG. 21. Referring to FIG. 21, in the center of the mask, as shown in FIG.
The four basic optical element patterns A to D shown in are formed by, for example, EB drawing. By forming a plurality of sets (three sets in FIG. 21) of the basic optical element patterns A and using each set on average, the effect of averaging the manufacturing error in the EB drawing can be obtained.

Further, three alignment marks 55 are drawn around the mask. The alignment mark 55 is used for alignment (positioning) when the diffractive optical element is formed into an integrated type according to the step-and-repeat method. Further, a pair of cutting guide patterns 56 is drawn on the mask. The pair of cutting guide patterns 56, together with the basic optical element patterns A to D, are exposed on the glass substrate. In this way, a plurality of diffractive optical elements are generated on the glass substrate, but the cutting guide pattern transferred to the glass substrate (cut line indicated by reference numeral 56a in FIG. 22).
The plurality of diffractive optical elements can be individually separated by cutting the substrate along the line.

Further, on the mask, for example, a line-and-space pattern 57 is formed as a control rule pattern for controlling the line width and the depth. Instead of the line and space pattern 57,
Dot patterns can be formed. The patterns for the diffractive optical element, that is, the basic optical element patterns A to D are
As described above, it is a complicated pattern in which a desired portion of the FZP pattern is divided and rearranged densely. Therefore,
The inside thereof is curved, and the boundary of the partial optical element has a discontinuous portion, and it is difficult to measure the line width of the transferred pattern. Therefore, in the first embodiment, the line width control line
The AND space pattern 57 is provided on the mask.

In FIG. 22, a dummy portion is formed on the left side of the cut line 56a, and a used portion is formed on the right side. A pattern 57a corresponding to the line-and-space pattern 57 of FIG. 21 is transferred to the dummy portion. On the other hand, in the used portion, one basic diffractive optical element pattern is formed by integrating the basic optical element patterns A to D of the mask by reduction projection exposure according to the step-and-repeat method. At the time of exposure, the shift amount s along the y direction is set between two sets of basic optical element patterns A to D that are adjacent to each other in the x direction.
y'is set. The exposure shift amount sy ′ is shown in FIG.
Based on the shift amount sy and the projection magnification mag of
= 2 · mag · sy

The line-and-space pattern 57a of the dummy portion measures the line width of the pattern of the diffractive optical element at the time of exposing on the resist, developing, etching with the resist, and removing the resist. However, it is used as a line width measurement point for mass production while confirming that the measured line width is within the error of the target line width. In addition, the line-and-space pattern 57a is used as a depth measurement point for achieving a desired etching depth by performing the etching step and the step of measuring the etched depth a plurality of times. After finally achieving the desired pitch (line width) and depth, this dummy part is unnecessary and is cut off by a cutting process. However, cutting is not essential in the present invention, and it is possible to leave the dummy portion as it is.

In the first embodiment, the projection basic area PIA1 obtained by projecting the ring-shaped basic illumination area IA1 to be formed on the incident surface of the fly-eye integrator 6.
Is divided into a large number of partial areas, which is equivalent to dividing the basic illumination area IA1 into a large number of partial illumination areas and then obtaining a large number of partial areas by projection. Also, the first
In the embodiment, the unit length Lx_rea is set larger than the length in the x direction of each partial area (see FIG. 13A) obtained by equally dividing the projection basic area PIA1 in the y direction. With, a design using a smaller unit length Lx_rea is also possible.

(Second Embodiment) Next, as a second embodiment, a structure of a diffractive optical element 22 for circular illumination used together with a circular aperture stop 32 in ordinary circular illumination and a manufacturing method thereof will be described. FIG. 23 is a diagram showing the relationship between a circular illumination region to be formed on the incident surface of the fly-eye integrator and a large number of lens elements in the second embodiment. Referring to FIG. 23, a part of a large number of lens elements constituting the fly-eye integrator 6 which are relevant to the description and the aperture stop 3 are shown.
2 shows a circular transmission area AP2 (shown by a broken line in the figure) and a circular basic illumination area IA2 (shown by a broken line in the figure) to be formed on the incident surface of the fly-eye integrator 6.

However, in FIG. 23, regarding the four lens elements near the center O of the fly-eye integrator 6, an unused area for cutting the 0th order light (straight light) generated from the diffractive optical element 22 is formed. is doing. However, for highly uniform illumination, the basic illumination area IA2 is designed to include this unused area. The rectangular area SP drawn inside the rectangular area of each lens element EL is an area obtained by proportionally reducing the rectangular area of each lens element EL, and the emission of the fly-eye integrator 6 is performed. It is a light spot region formed on the side.

In FIG. 23, it is assumed that the filling rate of each light spot region is 50% with respect to the rectangular region of each lens element EL. Here, the lens element EL whose light spot area SP is included in the transmission area AP2 of the aperture stop 32 is a necessary lens element, and FIG. 23 illustrates a part of the necessary lens element. On the other hand, the basic illumination area IA2
Is set by adding a margin portion to the outermost circumference of the aggregate of necessary lens elements. Similarly to the first embodiment, the setting of the necessary lens elements and the setting of the basic illumination area IA2 based on the necessary lens elements are finally set by simulation or the like so that the illumination uniformity is satisfied after including various tolerances. Is desirable.

FIG. 24 is a diagram showing a projected basic area obtained by projecting the basic illumination area onto a predetermined virtual diffraction lens in the second embodiment. Referring to FIG. 24, a projection basic area P obtained by projecting the basic illumination area IA2 on a virtual diffractive lens virtually arranged at the position of the diffractive optical element 22.
Of IA2, only the upper half divided projection basic region of 0 ≦ y is shown. As described above, the projection basic area PIA2
Since the virtual diffractive lens portion included in (1) corresponds to the basic optical element, the projection basic area PIA2 and the basic optical element are identified as the same. In the second embodiment, as in the first embodiment, a phase type binary Fresnel zone plate is used as the virtual diffractive lens.

FIG. 25 (a) is a diagram showing a state in which the upper half divided projection basic area is divided into a large number of partial areas in the second embodiment, and FIG. 25 (b) shows a plurality of divided partial areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.
In FIG. 25A, the upper half of the divided projection basic area PIA2
Is divided into 26 partial regions extending along the x direction. Here, each partial region has the same dimension along the y direction. That is, in the second embodiment, the divided projection basic area PIA2 is divided at equal intervals qy along the y direction.

Then, the total length Lx_to is obtained by arranging the partial regions divided at equal intervals along the y direction in a line along the x direction.
A value obtained by dividing t by the division number my = 14 is a unit length Lx_rea.
I am trying. Thus, the 26 partial regions shown in FIG. 25A are set by the same procedure as in the first embodiment. Then, the 26 partial regions shown in FIG. 25A and the Oz set at or near the origin O of FIG.
Overlap with the FZP pattern centered on
By dividing this FZP pattern according to the 26 partial regions, 26 partial optical elements are obtained. Figure 2
5 (b), the 26 partial optical elements defined corresponding to the 26 partial regions are densely rearranged, the dimension along the x direction is Lx_rea, and the dimension along the y direction is (1
A 4 × qy) rectangular divided basic optical element Aconv + is formed.

Similarly, the divided projection basic area (not shown) in the lower half of y <0 can be divided into 26 partial areas, and thus 26 partial optical elements corresponding to the 26 partial areas. Is obtained. Then, the 26 partial optical elements defined corresponding to the 26 partial regions are densely rearranged, and the dimension along the x direction is Lx_rea and the dimension along the y direction is (14 × qy). A rectangular divided basic optical element Aconv- (not shown) is formed.

FIG. 26 is a diagram showing how the basic optical element is formed by synthesizing two divided basic optical elements in the second embodiment. As shown in FIG. 26, the basic optical element Aconv for forming the entire circular basic illumination area IA2 as shown in FIG. 23 on the incident surface of the fly-eye integrator 6 is the same as the split basic element shown in FIG. Optical element Aconv + and split basic optical element Aco (not shown)
It can be obtained by arranging nv- and adjacent to each other along the y-direction. However, regarding one set of the divided basic optical element Aconv + and the divided basic optical element Aconv−,
Without being limited to this, various modifications are possible.

FIG. 27 is a diagram showing a state in which four types of basic optical elements in which the positions of the basic illumination areas are slightly different are arranged in the second embodiment. In FIG. 27, it is only the basic optical element pattern Aconv that is actually designed by decentering the center O of the projection illumination area PIA2 and the center Oz of the FZP pattern along the x direction and the y direction. Then, the basic optical element pattern Aconv_x
Is generated by subjecting the basic optical element pattern Aconv to front-back inversion in the x direction, that is, left-right inversion. The basic optical element pattern Aconv_y is generated by subjecting the basic optical element pattern Aconv to inversion in the y direction, that is, upside down. Basic optical element pattern Acon
v_xy is generated by vertically inverting the basic optical element pattern Aconv_x.

The above four types of basic optical element patterns are
It is formed on the mask according to the projection magnification mag. Then, the pattern of the diffractive optical element is formed by integrating the basic optical element pattern of the mask by projection exposure according to the step-and-repeat method. Also in the second embodiment, as shown in FIGS. 20A and 20B of the first embodiment, the number of overlaps of the basic optical element patterns is set to 9 or 16, or more overlaps are performed. It can be set to a number. It is desirable that the number of overlaps be determined based on illumination characteristics such as illumination unevenness and power. Further, in the manufacture of the diffractive optical element 22, the point that a simple dummy pattern such as a line and space pattern is used for controlling the line width of the transfer pattern and the etching depth is the same as in the first embodiment. is there.

(Third Embodiment) Next, as a third embodiment, a structure of a diffractive optical element 24 for quadrupole illumination used together with a quadrupole aperture stop 31 for quadrupole illumination and a method of manufacturing the same will be described. FIG. 28 is a diagram showing the relationship between the quadrupole aperture of the flyeye integrator and the quadrupole aperture stop in the third embodiment, and the quadrupole illumination area to be formed on the incident surface of the flyeye integrator. Is.

Referring to FIG. 28, a large number of lens elements EL constituting the fly-eye integrator 6 and four circular openings of the 4-pole aperture stop 31, that is, the transmission area A.
P3 and four circular sub-basic illumination areas IA3a and IA3 that form a quadrupole basic illumination area (shown by a broken line in the figure) to be formed on the incident surface of the fly-eye integrator 6.
b, IA3c and IA3d are shown. In addition,
The rectangular area SP drawn inside the rectangular area of each lens element EL is an area obtained by proportionally reducing the rectangular area of each lens element EL, and is located on the exit side of the fly-eye integrator 6. It is a light spot region to be formed. In FIG. 28, it is assumed that the filling rate of each light spot region is 50% with respect to the rectangular region of each lens element EL.

Here, the light spot region SP is the aperture stop 3
The lens element EL included in the four transmissive areas AP3 of 1 is a necessary lens element. On the other hand, the four sub-basic illumination areas IA3a to IA3d are set by adding a margin portion to the outermost periphery of the aggregate of necessary lens elements. Similar to the first and second embodiments, the setting of the necessary lens elements and the setting of the sub-basic illumination areas IA3a to IA3d based on the setting finally satisfy the illumination uniformity after including various tolerances. As described above, it is desirable to set it by simulation or the like. that time,
It is desirable to simultaneously comprehensively evaluate the contributions from the four sub-basic illumination areas IA3a-IA3d.

Further, in FIG. 28, the centers of the four sub-basic illumination areas IA3a to IA3d are indicated by reference signs a to d, respectively. Hereinafter, for simplicity of explanation, quadrupole illumination symmetrical about the optical axis will be considered. In this case, four sub-basic illumination areas IA3a forming a quadrupole-shaped basic illumination area
Since ~ IA3d is symmetrical with respect to the optical axis, the sub-basic illumination area IA3a in the upper left portion (area of x <0, 0 <y) in the figure.
The third embodiment will be described focusing on the above. In the case of quadrupole illumination asymmetrical with respect to the optical axis, it is necessary to individually design the four sub-basic illumination areas IA3a to IA3d.

FIG. 29 is a diagram showing a sub-projection basic area obtained by projecting the sub-basic illumination area on a predetermined virtual diffraction lens in the third embodiment. Referring to FIG. 29,
Of the sub-projection basic area PIA3a obtained by projecting the sub-basic illumination area IA3a onto the virtual diffractive lens virtually arranged at the position of the diffractive optical element 24, only the upper half divided projection basic area is shown. As described above, since the virtual diffractive lens portion included in the sub projection basic area PIA3a corresponds to the sub basic optical element, the sub projection basic area PI
The A3a and the sub-basic optical element are identified as the same. In the third embodiment as well, as in the first and second embodiments, a phase type binary Fresnel zone plate is used as the virtual diffraction lens.

FIG. 30 (a) is a diagram showing a state in which the upper half divided projection basic area is divided into a large number of partial areas in the third embodiment, and FIG. 30 (b) shows a plurality of divided partial areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.
In FIG. 30A, the upper half of the divided projection basic area PIA3
a is divided into 10 partial regions extending along the x direction. Here, each partial region has the same dimension along the y direction. That is, in the third embodiment, the divided projection basic area PIA3a is divided at equal intervals qy in the y direction.

Then, the total length Lx_to is obtained by arranging the partial regions divided at equal intervals along the y direction in a line along the x direction.
The unit length Lx_rea is a value obtained by dividing t by the division number my = 6. In this way, the 10 partial regions shown in FIG. 30A are set by the same procedure as in the first and second embodiments. Then, 10 shown in FIG.
By overlapping the FZP pattern centered on the origin O or the Oz set near the origin O in FIG. 30 (a) and dividing the FZP pattern in accordance with the 10 partial regions. Ten partial optical elements are obtained. In FIG. 30B, 10 partial optical elements defined corresponding to 10 partial regions are densely rearranged, the dimension along the x direction is Lx_rea, and the dimension along the y direction is ( A rectangular divided basic optical element A4pole_a + of 10 × qy) is formed.

Similarly, the lower half divided projection basic region (not shown) can be divided into 10 partial regions, and 10 partial optical elements corresponding to the 10 partial regions can be obtained. Then, the 10 partial optical elements defined corresponding to the 10 partial regions are densely rearranged, and the dimension along the x direction is Lx_rea and the dimension along the y direction is (6 × qy). Rectangular divided basic optical element A
4pole_a- (not shown) is formed.

FIG. 31 is a view showing a state in which a sub basic optical element is formed by combining two divided basic optical elements and a basic optical element is formed by combining four sub basic optical elements in the third embodiment. is there. As shown in FIG. 31A, the sub-basic optical element A4pole_a for forming the entire sub-basic illumination area IA3a shown in FIG. 28 on the incident surface of the fly-eye integrator 6 has
It is obtained by arranging the divided basic optical element A4pole_a + of 0 (b) and the divided basic optical element A4pole_a− (not shown) adjacently in the y direction. However, the divided basic optical element A4pole_a + and the divided basic optical element A
The one set with 4 pole_a- is not limited to this, and various modifications are possible.

Further, as shown in FIG. 31 (b), on the incident surface of the fly-eye integrator 6, four sub-basic illumination areas IA3a to IA3d shown in FIG. 28 are formed.
The basic optical element A4pole for forming the entire polar basic illumination area is the sub-basic optical element A shown in FIG.
4pole_a and three sub-basic optical elements A4 not shown
pole_b, A4pole_c and A4pole_
It is obtained by arranging d and two-dimensionally adjacently. However, four sub basic optical elements A4pole_a to A4
The one set of pole_d is not limited to this, and various modifications are possible.

FIG. 32 is a diagram showing the arrangement of four types of basic optical elements in which the positions of the basic illumination areas are slightly different in the third embodiment. In FIG. 32, the center O of the quadrupole projection illumination area composed of the four sub-projection illumination areas PIA3a to PIA3d and the center Oz of the FZP pattern are decentered along the x direction and the y direction for actual design. There is only the basic optical element pattern A4pole. Then, the basic optical element pattern A4pole
_X is generated by subjecting the basic optical element pattern A4pole to x-direction front-back inversion, that is, left-right inversion.
The basic optical element pattern A4pole_y is generated by subjecting the basic optical element pattern A4pole to y-direction front-back inversion, that is, vertical inversion. The basic optical element pattern A4pole_xy is the basic optical element pattern A4p.
It is generated by subjecting ole_x to upside down.

The above four types of basic optical element patterns are
It is formed on the mask according to the projection magnification mag. Then, the pattern of the diffractive optical element is formed by integrating the basic optical element pattern of the mask by projection exposure according to the step-and-repeat method. In the third embodiment as well, as shown in FIGS. 20A and 20B of the first embodiment, the number of overlaps of the basic optical element patterns is set to 9 or 16, or more overlaps are performed. It can be set to a number. It is desirable that the number of overlaps be determined based on illumination characteristics such as illumination unevenness and power. Further, in the manufacture of the diffractive optical element 24, the point that a simple dummy pattern such as a line and space pattern is used for controlling the line width of the transfer pattern and the etching depth is the same as in the first embodiment. is there.

Further, in the third embodiment, the 4-pole aperture stop has four circular openings, and four circular sub-basic illumination areas are set in accordance with the four circular openings. An example is described. However, the present invention is not limited to this, and even when the four openings of the 4-pole aperture stop have shapes other than the circular shape, 4
It is possible to set one sub-basic illumination area and design the same as in the third embodiment.

(Fourth Embodiment) Next, as a fourth embodiment, a structure of a diffractive optical element 25 for dipole illumination used together with an x-direction dipole aperture stop 37 for dipole illumination and a manufacturing method thereof will be described. . FIG. 33 shows the relationship between the fly-eye integrator in the fourth embodiment, the di-polar opening of the x-direction di-polar aperture stop, and the di-polar illumination area to be formed on the incident surface of the fly-eye integrator. FIG.

Referring to FIG. 33, a large number of lens elements EL constituting the fly-eye integrator 6, two circular openings of the x-direction two-pole aperture stop 37, that is, two transmission areas AP4, and a fly-eye Two circular sub-basic illumination areas IA forming a dipole-shaped basic illumination area (shown by a broken line in the figure) to be formed on the incident surface of the integrator 6.
4a and IA4b are shown. The rectangular area SP drawn inside the rectangular area of each lens element EL is an area obtained by proportionally reducing the rectangular area of each lens element EL, and the emission of the fly-eye integrator 6 is performed. It is a light spot region formed on the side. In FIG. 33, it is assumed that the filling rate of each light spot region is 50% with respect to the rectangular region of each lens element EL.

Here, the light spot region SP is the aperture stop 3
The lens element EL included in the two transmissive areas AP4 of 7 is a necessary lens element. On the other hand, the two sub-basic illumination areas IA4a and IA4b are set by adding a margin portion to the outermost circumference of the assembly of necessary lens elements. Similar to the first to third embodiments, the setting of the necessary lens elements and the setting of the sub-basic illumination areas IA4a and IA4b based on the necessary lens elements finally include various tolerances and the illumination uniformity is satisfied. As described above, it is desirable to set it by simulation or the like. In that case, it is desirable to simultaneously comprehensively evaluate the contributions from the two sub-basic illumination areas IA4a and IA4b.

Further, in FIG. 33, the centers of the two sub-basic illumination areas IA4a and IA4b are indicated by reference signs a and b, respectively. Hereinafter, for the sake of simplicity of description, consider a dipole illumination that is symmetrical with respect to the optical axis. in this case,
Since the two sub-basic illumination areas IA4a and IA4b forming the dipole-shaped basic illumination area are symmetrical with respect to the optical axis, the sub-basic illumination area IA on the left side (the area of x <0) in the figure.
4th Embodiment is described focusing on 4a. In the case of the two-pole illumination asymmetric with respect to the optical axis, it is necessary to individually design the two sub-basic illumination areas IA4a and IA4b.

FIG. 34 is a diagram showing a sub-projection basic area obtained by projecting the sub-basic illumination area on a predetermined virtual diffraction lens in the fourth embodiment. Referring to FIG. 34,
Of the sub-projection basic area PIA4a obtained by projecting the sub-basic illumination area IA4a onto the virtual diffractive lens virtually arranged at the position of the diffractive optical element 25, only the upper half divided projection basic area is shown. As described above, since the virtual diffractive lens portion included in the sub projection basic area PIA4a corresponds to the sub basic optical element, the sub projection basic area PI
The A4a and the sub-basic optical element are identified as the same. In the fourth embodiment, as in the first to third embodiments, the phase type binary Fresnel zone plate is used as the virtual diffractive lens.

FIG. 35 (a) is a diagram showing a state in which the upper half divided projection basic area in the fourth embodiment is divided into a large number of partial areas, and FIG. 35 (b) shows a plurality of divided partial areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.
In FIG. 35A, the upper half of the divided projection basic area PIA4
a is divided into 10 partial regions extending along the x direction. Here, each partial region has the same dimension along the y direction. That is, in the fourth embodiment, the divided projection basic area PIA4a is divided at equal intervals qy in the y direction.

Then, the total length Lx_to is obtained by arranging the partial regions divided at equal intervals along the y direction in a line along the x direction.
The unit length Lx_rea is a value obtained by dividing t by the division number my = 6. Thus, the 10 partial regions shown in FIG. 35A are set by the same procedure as in the first to third embodiments. Then, the ten partial regions shown in FIG. 35A and the FZP pattern centered on Oz set at or near the origin O in FIG. By dividing according to the partial area of, 10 partial optical elements can be obtained. In FIG. 35 (b), 10 partial optical elements defined corresponding to 10 partial regions are densely rearranged, the dimension along the x direction is Lx_rea, and the dimension along the y direction is ( A rectangular divided basic optical element A2pole_x_a + of 10 × qy) is formed.

Similarly, the divided projection basic area (not shown) in the lower half can also be divided into 10 partial areas, and 10 partial optical elements corresponding to the 10 partial areas can be obtained. Then, the 10 partial optical elements defined corresponding to the 10 partial regions are densely rearranged, and the dimension along the x direction is Lx_rea and the dimension along the y direction is (6 × qy). Rectangular divided basic optical element A
2pole_x_a- (not shown) is formed.

FIG. 36 is a view showing a state in which a sub basic optical element is formed by combining two divided basic optical elements and a basic optical element is formed by combining two sub basic optical elements in the fourth embodiment. is there. As shown in FIG. 36A, the sub-basic optical element A2pole_a for forming the entire sub-basic illumination area IA4a shown in FIG. 33 on the incident surface of the fly-eye integrator 6 has
This is obtained by arranging the divided basic optical element A2pole_x_a + of 5 (b) and the divided basic optical element A2pole_x_a- (not shown) adjacently in the y direction. However, one set of the divided basic optical element A2pole_x_a + and the divided basic optical element A2pole_x_a- is not limited to this, and various modifications are possible.

Further, as shown in FIG. 36 (b), the entire dipole-shaped basic illumination area consisting of the two sub-basic illumination areas IA4a and IA4b shown in FIG. 33 is formed on the incident surface of the fly-eye integrator 6. The basic optical element A2pole for doing so is the sub-basic optical element A2pole_a shown in FIG. 36A and the sub-basic optical element A2p (not shown).
It is obtained by disposing ole_b and ole_b adjacent to each other along the x direction. However, two sub-basic optical elements A2pol
The one set of e_a and A2pole_b is not limited to this, and various modifications are possible.

(Fifth Embodiment) Next, as a fifth embodiment, a structure of a diffractive optical element 26 for dipole illumination used together with a y-direction dipole aperture stop 38 in dipole illumination and a manufacturing method thereof will be described. . FIG. 37 shows the relationship between the fly-eye integrator in the fifth embodiment, the di-polar opening of the y-direction di-polar aperture stop, and the di-polar illumination area to be formed on the incident surface of the fly-eye integrator. FIG.

Referring to FIG. 37, a large number of lens elements EL constituting the fly-eye integrator 6, two circular openings of the y-direction two-pole aperture stop 37, that is, two transmission areas AP5, and a fly-eye Two circular sub-basic illumination areas IA forming a dipole-shaped basic illumination area (shown by a broken line in the figure) to be formed on the incident surface of the integrator 6.
5a and IA5b are shown. The rectangular area SP drawn inside the rectangular area of each lens element EL is an area obtained by proportionally reducing the rectangular area of each lens element EL, and the emission of the fly-eye integrator 6 is performed. It is a light spot region formed on the side. In FIG. 37, it is assumed that the filling rate of each light spot region is 50% with respect to the rectangular region of each lens element EL.

Here, the light spot region SP is the aperture stop 3
The lens element EL included in the two transmission areas AP5 of 8 is a necessary lens element. On the other hand, the two sub-basic illumination areas IA5a and IA5b are set by adding a margin portion to the outermost circumference of the assembly of necessary lens elements. Similar to the first to fourth embodiments, the setting of the necessary lens elements and the setting of the sub-basic illumination areas IA5a and IA5b based on the necessary lens elements finally include various tolerances and satisfy the illumination uniformity. As described above, it is desirable to set it by simulation or the like. In that case, it is desirable to comprehensively evaluate the contributions from the two sub-basic illumination areas IA5a and IA5b simultaneously.

Further, in FIG. 37, the centers of the two sub-basic illumination areas IA5a and IA5b are indicated by reference signs a and b, respectively. Hereinafter, for the sake of simplicity of description, consider a dipole illumination that is symmetrical with respect to the optical axis. in this case,
Since the two sub-basic illumination areas IA5a and IA5b forming the dipole-shaped basic illumination area are symmetrical with respect to the optical axis, the sub-basic illumination area IA on the left side (area of 0 <y) in the figure.
The fifth embodiment will be described by focusing on 5a. In the case of dipole illumination asymmetric with respect to the optical axis, it is necessary to individually design the two sub-basic illumination areas IA5a and IA5b.

FIG. 38 is a diagram showing a sub-projection basic area obtained by projecting the sub-basic illumination area on a predetermined virtual diffractive lens in the fifth embodiment. Referring to FIG. 38,
Of the sub-projection basic area PIA5a obtained by projecting the sub-basic illumination area IA5a onto the virtual diffractive lens virtually arranged at the position of the diffractive optical element 26, only the upper half divided projection basic area is shown. As described above, since the virtual diffractive lens portion included in the sub projection basic area PIA5a corresponds to the sub basic optical element, the sub projection basic area PI
The A5a and the sub basic optical element are identified as the same. In the fifth embodiment as well, similar to the first to fourth embodiments, a phase type binary Fresnel zone plate is used as the virtual diffraction lens.

FIG. 39 (a) is a diagram showing a state in which the upper half divided projection basic area is divided into a large number of partial areas in the fifth embodiment, and FIG. 39 (b) is a diagram showing a large number of divided partial areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.
In FIG. 39 (a), the upper half of the divided projection basic area PIA4
a is divided into 10 partial regions extending along the x direction. Here, each partial region has the same dimension along the y direction. That is, in the fifth embodiment, the divided projection basic area PIA5a is divided at equal intervals qy in the y direction.

Then, the total length Lx_to is obtained by arranging the partial regions divided at equal intervals along the y direction in a line along the x direction.
The unit length Lx_rea is a value obtained by dividing t by the division number my = 6. Thus, the ten partial regions shown in FIG. 39A are set by the same procedure as in the first to fourth embodiments. Then, the ten partial regions shown in FIG. 39A and the FZP pattern centered on Oz set at or near the origin O in FIG. 39A are overlapped, and ten FZP patterns are formed. By dividing according to the partial area of, 10 partial optical elements can be obtained. In FIG. 39 (b), 10 partial optical elements defined corresponding to 10 partial regions are densely rearranged, the dimension along the x direction is Lx_rea, and the dimension along the y direction is ( A rectangular divided basic optical element A2pole_y_a + of 10 × qy) is formed.

Similarly, the lower half divided projection basic region (not shown) can be divided into 10 partial regions, and 10 partial optical elements corresponding to the 10 partial regions can be obtained. Then, the 10 partial optical elements defined corresponding to the 10 partial regions are densely rearranged, and the dimension along the x direction is Lx_rea and the dimension along the y direction is (6 × qy). Rectangular divided basic optical element A
2pole_y_a- (not shown) is formed.

FIG. 40 is a diagram showing a manner in which a sub basic optical element is formed by combining two divided basic optical elements and a basic optical element is formed by combining two sub basic optical elements in the fifth embodiment. is there. As shown in FIG. 40A, the sub-basic optical element A2pole_a for forming the entire sub-basic illumination area IA5a shown in FIG. 37 on the incident surface of the fly-eye integrator 6 has
It is obtained by arranging the divided basic optical element A2pole_y_a + of 9 (b) and the divided basic optical element A2pole_y_a- (not shown) adjacently in the y direction. However, one set of the divided basic optical element A2pole_y_a + and the divided basic optical element A2pole_y_a- is not limited to this, and various modifications are possible.

Further, as shown in FIG. 40B, the entire dipole-shaped basic illumination area consisting of the two sub-basic illumination areas IA5a and IA5b shown in FIG. 37 is formed on the incident surface of the fly-eye integrator 6. The basic optical element A2pole for doing so is the sub-basic optical element A2pole_a shown in FIG.
It is obtained by disposing ole_b and ole_b adjacent to each other along the x direction. However, two sub-basic optical elements A2pol
The one set of e_a and A2pole_b is not limited to this, and various modifications are possible.

Although not shown, in the fourth and fifth embodiments as well, similar to the other embodiments, the number of overlaps of the basic optical element patterns is set to 4, 9 or 16, or more. It is possible to set the number of overlaps. It is desirable that the number of overlaps be determined based on illumination characteristics such as illumination unevenness and power. Further, in manufacturing the diffractive optical element 25 or 26, the point that a simple dummy pattern such as a line-and-space pattern is used for controlling the line width of the transfer pattern and the etching depth is the same as the first embodiment. It is the same.

Further, in the fourth and fifth embodiments, the two-pole aperture stop has two circular apertures, and two circular sub apertures are formed in accordance with the two circular apertures. An example of setting the basic illumination area is described. However, without being limited to this, even when the two openings of the two-pole aperture stop have a shape other than a circular shape, two sub-basic illumination areas are set according to the shape of the openings,
It is possible to design similarly to the fourth and fifth embodiments.

By the way, in the case of the binary (two-stage) diffractive optical elements used in the first to fifth embodiments, diffracted light is generated symmetrically with respect to the optical axis. , The area of the basic optical element pattern is 1
It is also possible to reduce it to / 2. Of course, for performance stability, it is desirable not to reduce the area as described in each of the embodiments, but in terms of cost, the smaller the area pattern, the better. Hereinafter, this point will be specifically described for each embodiment.

In the first embodiment, referring to the example of four times averaging shown in FIG. 18, due to the symmetry of the diffracted light, only the basic optical element pattern A has the basic optical element pattern D (= A_
A basic illumination area according to xy) can also be generated.
Further, it is possible to generate a basic illumination area by the basic optical element pattern C (= A_y) only with the basic optical element pattern B. Therefore, even if the basic optical element patterns C and D are omitted, the effect of averaging four times can be roughly obtained. However, as described above, it is desirable not to omit the basic optical element patterns C and D for performance stability.

Similarly, referring to the example of four times averaging shown in FIG. 27 in the second embodiment, the basic optical element pattern A
If conv_y and Aconv_xy are omitted, 4
It is possible to roughly obtain the effect of the time average. However, in this case as well, for performance stability, it is desirable not to omit the basic optical element patterns Aconv_y and Aconv_xy. Similarly, referring to the example of four-time averaging shown in FIG. 32 in the third embodiment, the effect of four-time averaging can be roughly obtained even if the basic optical element patterns A4pole_y and A4pole_xy are omitted. However, also in this case, in order to stabilize the performance, the basic optical element patterns A4pole_y and A4 are used.
It is desirable not to omit pole_xy.

Similarly, although not shown, in the fourth and fifth embodiments, the basic optical element pattern A is used.
Even if 2pole_y and A2pole_xy are omitted, the effect obtained by averaging four times can be roughly obtained. However, also in this case, in order to stabilize the performance, the basic optical element patterns A2pole_y and A2pole_y
It is desirable not to omit _xy. Although the case of four-time averaging has been described as an example, even in the case of multiple-time averaging larger than 4, it is possible to omit the basic optical element pattern having symmetry about the optical axis.

(Sixth Embodiment) Next, as a sixth embodiment, the structure of a diffractive optical element for annular illumination used for annular illumination and the method for manufacturing the same will be described as in the first embodiment. However, the sixth embodiment is an embodiment capable of obtaining a more uniform illuminance distribution in the wafer conjugate plane and the illumination pupil plane than in the first embodiment. The sixth embodiment will be described below, focusing on the differences from the first embodiment.

FIG. 41 (a) is a diagram showing a state in which the upper half of the divided projection basic area is divided into a large number of partial areas in the sixth embodiment, and FIG. 41 (b) shows a plurality of divided partial areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.
As described above, FIG. 41 is a diagram corresponding to FIG. 13 in the first embodiment, but in FIG. 13A, the upper half divided projection basic area PIA1 is divided into 51 partial areas. On the other hand, in FIG.
It is further subdivided into 6 partial areas.

In the subdivision, the partial regions are divided so that the dimensions along the x direction are as equal as possible, the partial regions that can be equally divided are defined as equal divided partial regions, and the remaining partial regions that cannot be equally divided are divided. It is a non-equal-divided partial area. Then, 216 partial optical elements are defined corresponding to the 216 partial regions. Here, the partial optical element defined corresponding to the equally divided partial area is an even divided partial optical element, and the partial optical element defined corresponding to the unequal divided partial area is a unequal divided partial optical element. . For example, in FIG.
The partial optical element 1 in (a) is an equal division partial optical element 1 to 5 and a non-equal division partial optical element 6 in FIG.
It is divided into and.

Next, in the sixth embodiment, the 216 partial regions obtained by the above subdivision are randomly arranged.
Specifically, random replacement is performed in each column (partial optical element quantization line) extending along the x direction. FIG. 41 (a) shows the result of subdivision and random array in each column. At this point, 216
Only the indices (1-216) of the partial areas are randomly designated, and the step of defining the 216 partial optical elements by overlapping the FZP pattern on the divided projection basic area PIA1 is not performed.

In the sixth embodiment, in parallel with the subdivision and the random arrangement, the 51 partial optical elements densely arranged in FIG. 13B of the first embodiment are further subdivided into 216 partial optical elements. Turn into. Here, the rule of subdividing 51 partial optical elements into 216 partial optical elements in FIG. 41B is to subdivide 51 partial regions into 216 partial regions in FIG. 41A. It's all about the rules. For example, the partial optical element 1 in FIG.
41 is divided into equally divided partial optical elements 1 to 5 and non-uniformly divided partial optical element 6 in FIG.

Next, in the sixth embodiment, 216 partial optical elements obtained by the above subdivision are randomly arranged. Specifically, in FIG. 41B, the columns are randomly replaced in units of each column (rearrangement line) extending along the x direction. Further, the partial optical elements are randomly replaced in each row extending along the x direction. Further, the equally divided partial optical elements are randomly exchanged between the two rows extending along the x direction. In FIG. 41 (b), subdivision, random array for each column, random array in each column,
And the results of the random replacement of the equally divided partial optical elements between the columns are shown.

Thus, in the sixth embodiment, the 216 partial regions of FIG. 41 (a), which are subdivided and randomly arranged, and the origin O of FIG. 41 (a) or Oz set near the origin O are centered. 216 partial optical elements can be obtained by overlapping the FZP pattern with each other and dividing the FZP pattern according to the 216 partial regions. The 216 partial optical elements thus obtained are shown in FIG.
They are rearranged densely according to the random arrangement shown in FIG. 1 (b) to form a rectangular divided basic optical element P + (1).

Note that, in the above description, in the random arrangement in FIG. 41 (a), only random replacement within each column is performed, but random replacement of equally divided partial areas between two columns is performed. This makes it possible to obtain a better random array. In addition, FIG.
If the randomization in (b) is sufficient, then FIG.
An embodiment in which only subdivision is performed without performing the random arrangement in (a) is also possible. Further, when the randomization in FIG. 41 (a) is sufficient, an embodiment in which only the subdivision is performed without performing the random arrangement in FIG. 41 (b) is possible. However, in order to obtain a good random sequence, the random sequence in FIG. 41 (a) and the random sequence in FIG.
It is preferable to perform both of the random arrangements described above.

FIG. 42 (a) is a diagram showing a state in which the lower half divided projection basic area is divided into a large number of partial areas in the sixth embodiment, and FIG. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.
As shown in FIG. 42 (a), the lower half divided projection basic region of y <0 can be similarly divided into 216 partial regions, and thus 216 partial regions corresponding to the 216 partial regions. An optical element is obtained. And FIG.
As shown in (b), 216 partial optical elements defined corresponding to 216 partial regions are densely rearranged,
A rectangular divided basic optical element P- (1) is formed.

FIG. 43 is a diagram showing how the basic optical element is formed by synthesizing the basic divided optical element shown in FIG. 41 (b) and the basic divided optical element shown in FIG. 42 (b). A basic optical element P for forming the entire ring-shaped basic illumination area IA1 on the incident surface of the fly-eye integrator 6.
(1) is the divided basic optical element P + (1) of FIG. 41 (b)
And the divided basic optical element P- (1) in FIG. 42B are arranged adjacent to each other in the y direction. However,
Split basic optical element P + (1) and split basic optical element P-
The one set with (1) is not limited to this, and various modifications are possible.

In FIG. 43, P is a general notation.
Although (j) is used, j is the number of forms of the random array. That is, P (1) represents a basic optical element obtained through randomization according to the first mode, and P (j) represents a basic optical element obtained through randomization according to the j-th mode. Show. As will be described later, by changing the form of the random array, the above-mentioned operation is performed a plurality of times to form a plurality of basic optical elements according to different random arrays, and these are integrated on the diffractive optical element.

FIG. 44 is a diagram showing a state in which four types of basic optical elements in which the positions of the basic illumination areas are slightly different are arranged in the sixth embodiment. In FIG. 44, only the basic optical element pattern P (j) is actually designed by decentering the center O of the projection illumination area PIA1 and the center Oz of the FZP pattern along the x direction and the y direction. . The basic optical element pattern P (j) _x is generated by subjecting the basic optical element pattern P (j) to inversion in the x direction, that is, left / right inversion. The basic optical element pattern P (j) _y is the basic optical element pattern P (j)
Is generated by performing the up-down inversion in the y direction, that is, upside down. The basic optical element pattern P (j) _xy is generated by vertically inverting the basic optical element pattern P (j) _x.

In the sixth embodiment, four types of basic optical elements P (j), P (j) _x, P (j) _y and P (j) are used.
The aggregate of _xy is defined as a basic optical element aggregate AP (j). As described above, by using the basic optical element aggregate AP (j), it is possible to overlap the four basic illumination areas that are slightly different in position and perform illumination with improved uniformity. The basic optical element assembly obtained by repeating the above design 14 times for j = 1 to 14 is AP (1) to
This is AP (14). That is, for j = 2 to 14, the basic optical element P (j) having a random arrangement different in form from the random arrangement of j = 1 shown in FIGS. 41 and 42 is used.
To generate the basic optical element aggregate AP by applying the one-setting rule of FIG.
(J) is generated. Hereinafter, AP (j) will be referred to as APj.

FIG. 45 is a diagram schematically showing a pattern formed on a mask used for manufacturing the diffractive optical element in the sixth embodiment by lithography. Referring to FIG. 45, in the sixth embodiment, four block patterns AAP1 to AAP4 are formed. Here, the block pattern AAP1 is generated by arranging five basic optical element assemblies AP1 to AP5 adjacent to each other along the x direction.

The block pattern AAP2 is generated by arranging two basic optical element assemblies AP6 and AP7 adjacent to each other along the x direction. The block pattern AAP3 is generated by arranging two basic optical element assemblies AP8 and AP9 adjacent to each other in the x direction. The block pattern AAP4 is generated by arranging five basic optical element assemblies AP10 to AP14 adjacent to each other along the x direction.

FIG. 46 is a diagram schematically showing the structure of a mask used for manufacturing the diffractive optical element in the sixth embodiment by lithography. In addition, FIG.
It is a figure which shows one diffractive optical element of the some diffractive optical element produced | generated on the glass substrate using the mask of FIG. Referring to FIG. 46, in the center of the mask, as shown in FIG.
The four block patterns AAP1 to AAP4 shown in
For example, it is formed by EB drawing. Further, three alignment marks 65 are drawn around the mask. Here, the three alignment marks 65 are
It has the same function as the three alignment marks 55 in the first embodiment.

Further, a pair of cutting guide patterns 66 are drawn on the mask. Here, the pair of cutting guide patterns 66 has the same function as the pair of cutting guide patterns 56 in the first embodiment. Also,
On the mask, for example, a line-and-space pattern 67 is formed as a control rule pattern for controlling the line width and the depth. Here, the line and space pattern 67 has the same function as the line and space pattern 57 in the first embodiment.

Referring to FIG. 47, the four block patterns AAP1 to AAP4 of the mask are integrated by projection exposure according to the step-and-repeat method to form one diffractive optical element pattern. Specifically, a transfer pattern composed of four block patterns AAP1 to AAP4 is formed along the horizontal direction in the drawing. The transfer patterns formed of the same block pattern are staggered in the vertical direction in the drawing so that they are not adjacent to each other in the vertical direction in the drawing.

Such projection exposure can be performed in accordance with an exposure program for exposing the mask except for the necessary portions of the mask while shielding the light. Note that, in the manufacture of the diffractive optical element, the point that a simple dummy pattern such as a line and space pattern is used for controlling the line width of the transfer pattern and the etching depth is the same as in the first embodiment. . Also, since the dummy part to which the line and space pattern 67 is transferred is unnecessary, it can be separated by a cutting process, or the dummy part can be left as it is, similar to the first embodiment. .

Further, as described above, the binary (2
In the case of a (stage) diffractive optical element, diffracted light is generated symmetrically with respect to the optical axis. Therefore, it is possible to reduce the area of the basic optical element pattern to 1/2 by averaging design utilizing symmetry. Specifically, referring to the example of four times averaging shown in FIG. 44, due to the symmetry of the diffracted light, the basic optical element pattern P
It is possible to generate the basic illumination area by the basic optical element pattern P (j) _xy only with (j). Further, it is possible to generate a basic illumination area by the basic optical element pattern P (j) _y only with the basic optical element pattern P (j) _x. Therefore, the basic optical element pattern P (j)
Even if _y and P (j) _xy are omitted, the effect of four times average can be roughly obtained. However, as described above, it is desirable not to omit the basic optical element patterns P (j) _y and P (j) _xy for performance stability.

In the above description, the sixth embodiment has been described by setting the number of forms of the random array j = 14 and setting the number of subdivisions in the half area of the basic illumination area to 216. An embodiment in which the number of subdivisions is changed for each form is also possible. In this case, better randomization is possible than when the number of subdivisions is fixed. In the sixth embodiment, the partial optical elements are subdivided and randomized, and a large number of random combinations are performed.
Irregularities due to diffraction and manufacturing errors are averaged, and the illumination uniformity on the wafer conjugate plane and the illumination uniformity (σ uniformity or pupil distribution uniformity) on the illumination pupil are significantly improved. In the above example, the case of the ring illumination is described, but the cases of the normal circular illumination, the four-pole illumination, and the two-pole illumination described in the second to fifth embodiments are the same as those of the present embodiment. It is possible to carry out the same design as that of the present embodiment by performing similar subdivision, randomization, and the like.

(Seventh Embodiment) Next, as a seventh embodiment, the structure of a diffractive optical element for annular illumination used for annular illumination and the manufacturing method thereof will be described as in the first embodiment. However, in the seventh embodiment, the first
It is an embodiment capable of forming a ring-shaped basic illumination area having a boundary line which is much more circular than the embodiment. In recent years, in an exposure apparatus, a method aiming at power up by an illumination system that does not use an aperture stop (shown by reference numeral 33 in FIG. 7) is desired. In that case, since the divergence characteristic itself of the diffractive optical element determines the pupil shape, the desired pupil shape itself is not the shape in which the margin is added to the aggregate of necessary lens elements as in the first to sixth embodiments. It is necessary to form a basic illumination area having a shape approximately equal to.

That is, in an apertureless illumination optical system that does not use an aperture stop, it is desirable to form a circular basic illumination region having a boundary line that is as close to a circle as possible in normal circular illumination, and to the extent possible in annular illumination. It is desirable to form a ring-shaped basic illumination area with a border that is close to a circle. A diaphragmless illumination optical system that does not use an aperture diaphragm can be realized, for example, by using a microlens array instead of the fly-eye integrator 6.

In general, the microlens array is formed by, for example, etching a parallel flat glass plate to form a group of microlenses. Here, each microlens forming the microlens array is smaller than each lens element forming the fly-eye integrator. Further, unlike the fly-eye integrator that includes lens elements that are isolated from each other, the microlens array is integrally formed with a large number of microlenses that are not isolated from each other. However, the microlens array is the same as the fly-eye integrator in that the lens elements are arranged vertically and horizontally, and they both constitute a wavefront division type optical integrator. Hereinafter, in the seventh embodiment, a microlens array (not shown) is used instead of the fly-eye integrator 6.

FIG. 48 is a diagram showing the relationship between an annular illumination region to be formed on the entrance surface of the microlens array and a large number of lens elements in the seventh embodiment. Referring to FIG. 48, a part of a large number of minute lenses forming the microlens array, which are relevant to the description, and a ring-shaped basic illumination area (indicated by a broken line in the figure) to be formed on the incident surface of the microlens array. Shown) IA300 is shown. Here, the basic illumination area IA300 is set so as to have a boundary line that is as circular as possible along the shape of the ring-shaped virtual transmission area IAP of the virtual ring aperture stop.

Specifically, the basic illumination area IA300 is
The half region (0 ≦ y) of the ring-shaped virtual transmission region IAP is quantized in 43 steps along the y direction, so that the shape is approximated to the shape of the ring-shaped virtual transmission region IAP. By further increasing the quantization number, it is possible to set a ring-shaped basic illumination area that is further approximated to the shape of the ring-shaped virtual transmission area IAP. In the seventh embodiment, the ring-shaped projected basic illumination area P is obtained by projecting the ring-shaped basic illumination area IA300 obtained by 43-step quantization on a predetermined virtual diffraction lens.
An IA300 (not shown) is obtained.

FIG. 49 (a) is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions in the seventh embodiment, and FIG. 49 (b) is a diagram showing a plurality of divided partial regions. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.
In FIG. 49A, the upper half of the divided projection basic area PIA3
00 is divided into 98 partial regions extending along the x direction. Here, each partial region has the same dimension along the y direction. That is, in the seventh embodiment, the divided projection basic area PIA300 is divided at equal intervals qy in the y direction.

Then, the total length Lx_to is obtained by arranging the partial regions divided at equal intervals along the y direction in a line along the x direction.
A value obtained by dividing t by the division number my = 29 is a unit length Lx_rea.
I am trying. Thus, the 98 partial regions shown in FIG. 49A are set by the same procedure as in the first embodiment. Then, the 98 partial regions shown in FIG.
By overlapping the ZP pattern and dividing this FZP pattern according to 98 partial regions,
98 partial optical elements are obtained. In FIG. 49 (b),
The 98 partial optical elements defined corresponding to the 98 partial regions are densely rearranged, and the dimension along the x direction is Lx.
With _rea, a rectangular split elementary optical element A + having a dimension along the y direction of (29 × qy) is formed.

Similarly, the lower half divided projection basic area (not shown) of y <0 can be divided into 98 partial areas, and thus 98 partial optical elements corresponding to the 98 partial areas. Is obtained. Then, the 98 partial optical elements defined corresponding to the 98 partial regions are densely rearranged, and the dimension along the x direction is Lx_rea and the dimension along the y direction is (29 × qy). A rectangular divided basic optical element A- (not shown) is formed.

FIG. 50 is a diagram showing how a basic optical element is formed by combining two divided basic optical elements in the seventh embodiment. As shown in FIG. 50, the basic optical element A for forming the entire ring-shaped basic illumination area IA300 as shown in FIG. It is obtained by arranging A + and a divided basic optical element A- (not shown) adjacently along the y direction. However, one set of the divided basic optical element A + and the divided basic optical element A- is not limited to this, and various modified examples are possible.

Although not shown, in the seventh embodiment as well as in the first embodiment, the positions of the basic illumination regions are slightly different in order to reduce uneven illumination on the incident surface of the microlens array. Four types of basic optical elements can be arranged (see FIG. 18). In this case, only the basic optical element pattern A is actually designed by decentering the center O of the projection illumination area PIA300 and the center Oz of the FZP pattern along the x direction and the y direction.
Then, the basic optical element pattern B, that is, A_x, is generated by subjecting the basic optical element pattern A to inversion in the x direction, that is, left-right inversion. The basic optical element pattern C, that is, A_y, is generated by subjecting the basic optical element pattern A to face-to-face inversion in the y direction, that is, upside down. The basic optical element pattern D, A_xy, is generated by vertically inverting the basic optical element pattern B, A_x.

The basic optical element patterns A to D are formed on the mask according to the projection magnification mag. Then, the patterns of the diffractive optical element are formed by integrating the basic optical element patterns A to D of the mask by projection exposure according to the step-and-repeat method. In the manufacture of the diffractive optical element, the use of a simple dummy pattern such as a line and space pattern for controlling the line width of the transfer pattern and the etching depth is the same as in the first embodiment.

Also in the seventh embodiment, as shown in FIGS. 20A and 20B of the first embodiment, the basic optical element pattern overlap number is set to 9 or 16, or
Alternatively, it is possible to set the number of overlaps more than that. It is desirable that the number of overlaps be determined based on illumination characteristics such as illumination unevenness and power. Also,

As in the sixth embodiment, the partial optical elements can be further subdivided and randomized, and a large number of random combinations can be performed. in this case,
Also in the seventh embodiment, the unevenness of diffraction and the unevenness of illumination due to manufacturing errors are averaged, and the illumination uniformity on the wafer conjugate plane and the illumination uniformity (σ uniformity or pupil distribution uniformity) on the illumination pupil are significantly improved. .

(Eighth Embodiment) Next, as an eighth embodiment, similar to the seventh embodiment, the structure of a diffractive optical element for annular illumination used for annular illumination in a diaphragmless illumination system and its manufacture. The method will be described. However, the eighth embodiment is an embodiment capable of further improving the circular approximation of the boundary line that defines the ring-shaped basic illumination area, as compared with the seventh embodiment. The eighth embodiment will be described below, focusing on the differences from the seventh embodiment.

Referring to FIGS. 48 to 50 in the seventh embodiment, since the quantization of the partial region and the quantization of the partial optical element are performed along the y direction, the boundary line is large in the region where the absolute value of the y coordinate is large. The circular approximation of is slightly worse. In the eighth embodiment, in order to improve the circular approximation of the boundary line that defines the ring-shaped basic illumination area, the basic optical element designed by the quantization along the x direction and the quantization along the y direction are used. Almost the same number of basic optical elements to be designed are mixed. As a result, in the eighth embodiment, illumination without quantization anisotropy can be implemented.

FIG. 51 (a) is a diagram showing a state in which the right half divided projection basic area is divided into a large number of partial areas in the eighth embodiment, and FIG. 51 (b) shows a large number of divided partial areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.
In FIG. 51A, the right half (0 ≦ x) divided projection basic area PIA300 ′ (divided projection basic area PIA3 in FIG. 49).
00 obtained by rotating 90 in the xy plane by 90 degrees) is divided into 98 partial regions extending along the y direction.
Here, each partial region has the same dimension along the x direction. That is, in the eighth embodiment, the divided projection basic region P
The IA 300 'is divided along the x direction at equal intervals qx.

Then, the total length Ly_to in which partial regions divided at equal intervals along the x direction are arranged in a line along the y direction
A value obtained by dividing t by the number of divisions mx = 29 is a unit length Ly_rea.
I am trying. Thus, the 98 partial regions shown in FIG. 51A are set by the same procedure as in the first embodiment. Then, the 98 partial regions shown in FIG.
By overlapping the ZP pattern and dividing this FZP pattern according to 98 partial regions,
98 partial optical elements are obtained. In FIG. 51 (b),
The 98 partial optical elements defined corresponding to the 98 partial regions are densely rearranged, and the dimension along the y direction is Ly.
With _rea, a rectangular split elementary optical element A ′ + having a dimension (29 × qx) along the x direction is formed.

Similarly, the left half divided projection basic area (not shown) where x <0 can be divided into 98 partial areas, and thus 98 partial optical elements corresponding to the 98 partial areas. Is obtained. Then, the 98 partial optical elements defined corresponding to the 98 partial regions are densely rearranged, and the dimension along the y direction is Ly_rea and the dimension along the x direction is (29 × qx). A rectangular divided basic optical element A'- (not shown) is formed.

FIG. 52 is a diagram showing how the basic optical element is formed by combining two divided basic optical elements in the eighth embodiment. As shown in FIG. 52, a ring-shaped basic illumination area IA is formed on the incident surface of the microlens array.
The basic optical element A ′ for forming the entire 300 ′ (obtained by rotating the basic illumination area IA300 of FIG. 48 by 90 degrees in the xy plane) is the divided basic optical element A ′ + of FIG. 51 (b). And a divided basic optical element A'- (not shown) are arranged adjacent to each other along the x direction. However, one set of the divided basic optical element A ′ + and the divided basic optical element A′− is not limited to this, and various modifications are possible.

In the eighth embodiment, the basic optical element pattern A ′ quantized and designed in the x direction and the basic optical element pattern A quantized and designed in the y direction are set according to the projection magnification mag. Formed on the mask. Then, the patterns of the diffractive optical element are formed by integrating the basic optical element patterns A and A ′ of the mask by projection exposure according to the step-and-repeat method. That is, a mask including both the basic optical element patterns A and A ′ is manufactured. Then, using this mask, projection exposure is performed on the glass substrate so that the transfer patterns of the basic optical element pattern A and the transfer patterns of the basic optical element pattern A ′ are included in substantially the same number. Thus
A diffractive optical element is manufactured through a development and etching process.

FIG. 53 is a diagram showing one of a plurality of diffractive optical elements formed on a glass substrate in the eighth embodiment. Referring to FIG. 53, one basic diffractive optical element pattern is formed by integrating two basic optical element patterns A and A ′ by projection exposure according to a step-and-repeat method. Specifically, the transfer patterns of the basic optical element pattern A are continuously formed so as to be adjacent to each other in the y direction. Further, the transfer pattern by the basic optical element pattern A ′ is also continuously formed so as to be adjacent to each other along the y direction. Then, the transfer pattern row of the basic optical element pattern A and the transfer pattern row of the basic optical element pattern A ′ are formed so as to be adjacent to each other in the x direction.

As described above, in the eighth embodiment, the ring-shaped basic illumination area IA300 by the basic optical element A and the ring-shaped basic illumination area IA300 ′ by the basic optical element A ′ are overlapped and the averaged intensity distribution is obtained. , On the desired target surface (eg, the entrance surface of the microlens array). This averaged intensity distribution contains both the components related to the quantization in the x direction and the components related to the quantization in the y direction with almost the same weight, so that good illumination having no anisotropy in the x direction and the y direction is obtained. It can be carried out.

Referring to FIG. 53, since the basic optical element patterns A and A ′ are generally not square, the transfer pattern row by the basic optical element pattern A and the transfer pattern row by the basic optical element pattern A ′ are respectively formed in the y direction. It is formed along. As a result, it is possible to arrange a large number of transfer patterns of the basic optical element pattern A and a large number of transfer patterns of the basic optical element pattern A ′ without any gap.

However, although the aspect ratios of the basic optical element patterns A and A'are exaggerated in FIG. 53, a design close to a square is actually possible. Also, so that the basic optical element patterns A and A ′ are completely square,
It is also possible to set the projection basic areas PIA300 and 300 '. In this case, various modified examples of the transfer pattern by the basic optical element pattern A and the transfer pattern by the basic optical element pattern A ′ are possible.
As an example, it is also possible to arrange the transfer pattern of the basic optical element pattern A and the transfer pattern of the basic optical element pattern A ′ in a staggered or checkered pattern.

Further, similar to the sixth embodiment, in designing the basic optical element patterns A and A ′, the partial optical elements can be further subdivided and randomized, and a large number of random combinations can be performed. In this case, also in the eighth embodiment, the unevenness of diffraction and the unevenness of illumination due to manufacturing errors are averaged, and the illumination uniformity on the wafer conjugate plane and the illumination uniformity (σ uniformity or pupil distribution uniformity) on the illumination pupil are significantly improved. Improve to. Further, in the seventh embodiment and the eighth embodiment, the case of the ring-shaped illumination has been described, but the same design as that of the present embodiment is implemented also in the case of the normal circular illumination, the 4-pole illumination, and the 2-pole illumination. It is possible.

FIG. 54 is a view for explaining the first modification example of the second embodiment. In the second embodiment, the circular illumination area formed on the incident surface of the fly-eye integrator 6 realizes a uniform intensity distribution as shown by the broken line 70. However, depending on the case, it is required to form a circular illumination area in which the intensity distribution linearly gradually decreases in the periphery as shown by a broken line 71. Therefore, the second
In the first modification example of the embodiment, as shown in FIG. 54A, by combining a plurality of basic optical elements having different focal lengths, the intensity distribution gradually decreases in the periphery as shown by the solid line 72. Such a circular illumination area is formed.

In this case, the intensity distribution shown by the solid line 72 can be made closer to the desired intensity distribution shown by the broken line 71 by increasing the number of types of basic optical elements. In FIG. 54, for the sake of clarity of explanation, three types of basic optical elements f1 to f3 having different focal lengths are set. Here, each of the basic optical elements f1 to f3 is designed similarly according to the second embodiment, but the focal lengths thereof are set to be different from each other. And three types of basic optical elements f1 to f3
Are arranged as shown in FIG. 54 (b) to form an optical diffraction element. The arrangement shown in FIG. 54 (b) is advantageous when the incident light flux has a Gaussian intensity distribution along the vertical direction in the drawing and a top hat intensity distribution along the horizontal direction in the drawing. It is an array.

Although detailed description is omitted, the annular illumination in the first and sixth to eighth embodiments, the quadrupole illumination in the third embodiment, the fourth and fifth embodiments. Also in the bipolar illumination in the embodiment, as in the above-described modification, a plurality of basic optical elements having different focal lengths are combined to form an illumination area having a desired shape such that the intensity distribution gradually decreases in the periphery. be able to.

Incidentally, for example, in the first embodiment, the coordinate data of each partial optical element (for example, the coordinate data of four vertices) divided in FIG. 13A and the shape data of each partial optical element (divided FZP) are used. Pattern shape data). Here, the shape data of each partial optical element is expressed by a predetermined pattern depending on the FZP pattern. Then, based on the coordinate data of each partial optical element divided in FIG. 13A and the coordinate data of each partial optical element densely arranged in FIG. 13B, coordinate conversion of each partial optical element is performed. Get the data. And
Based on this coordinate conversion data, the shape data of each partial optical element is subjected to coordinate conversion. Here, the coordinate conversion of the shape data of each partial optical element is the coordinate conversion of a predetermined pattern.

According to a specific example, the large number of partial optical elements divided in FIG. 13A are mesh-divided into a larger number of cells. Here, the size of each cell optically corresponds to the size of each pixel in EB drawing, for example. Then, based on the large number of cells obtained by the mesh division and the predetermined pattern of each partial optical element, the cells that should form part of the pattern and the cells that do not form part of the pattern are determined. Thus, the coordinate conversion of the shape data of each partial optical element is the coordinate conversion of each cell information. The method of assuming this cell is CAD
Is suitable for a design method using.

In the seventh and eighth embodiments, a microlens array is used instead of the fly-eye integrator 6 to realize a diaphragmless illumination optical system that does not use an aperture diaphragm. A second modification in which an apertureless illumination optical system is realized by using an eye integrator is also possible. FIG. 55 is a diagram illustrating a configuration of a main part of a second modified example according to the seventh embodiment and the eighth embodiment.

In the second modification, as shown in FIG. 55, between the fly-eye integrator 6 and the diffractive optical element 80 (corresponding to the diffractive optical element 21 or the like in FIG. 7) according to the present invention, as shown in FIG. A relay optical system 81 having a function similar to that of the relay lens system 4 is provided. However, in FIG. 7, the front focal plane of the relay lens system 4 is aligned with the diffractive optical element 21, and the rear focal plane of the relay lens system 4 is aligned with the incident surface of the fly-eye integrator 6. In contrast, in FIG. 55, the front focal plane of the relay optical system 81 and the diffractive surface of the diffractive optical element 80 coincide with each other, and the rear focal plane of the relay optical system 81 and the exit surface position of the fly-eye integrator 6 are arranged. Are arranged so that and match.

In the second modification, the exit surface and the exit surface position are clearly distinguished. In FIG. 55, the diffractive optical element 80 and the exit surface of the fly-eye integrator 6 are conjugate, and the diffractive optical element 80 forms an image on the exit surface of the fly-eye integrator 6. On the other hand, the position of the exit surface of the fly-eye integrator 6 is set to the relay optical system 8
The light flux that is in the rear focal position of 1 but is incident parallel to the relay optical system 81 is reflected by the fly-eye integrator 6
Due to the refracting action on the incident surface of, the light is focused not on the exit surface of the fly-eye integrator 6 but on its center. That is, in the example of FIG. 55, the exit surface 6a of the fly-eye integrator 6 itself is not at the rear focal point of the relay optical system 81, but the fly-eye integrator 6 is not.
The exit surface position of is in the rear focus position of the relay optical system 81.

The light flux that has passed through the fly-eye integrator 6 passes through a condenser lens system 82 having a function similar to that of the condenser lens system 8 in FIG.
(Mask) 9 is illuminated in a superimposed manner. Here, as shown in FIG. 55, the front focal plane of the condenser lens system 82 and the incident plane position of the fly-eye integrator 6 coincide with each other, and the rear focal plane of the condenser lens system 82 and the reticle 9 are aligned.
Are arranged so that and match. The incident surface of the fly-eye integrator 6 and the reticle 9 are arranged optically conjugate with each other via the condenser lens system 82.

In the second modification, the incident surface and the incident surface position are clearly distinguished. When the parallel light flux shown by the broken line in FIG. 55 is traced backward from the reticle 9 side, this parallel light flux is not reflected on the exit surface of the fly-eye integrator 6 but is incident on the center of the fly-eye integrator 6 instead of the incident surface. Collect light. That is, in the example of FIG. 55, the incident surface 6b of the fly-eye integrator 6 is not at the front focal point of the condenser lens system 7, but the incident surface of the fly-eye integrator 6 is at the front focal point of the condenser lens system 7. In position.

Therefore, as shown by the solid line portion 6a in FIG. 55, a predetermined light intensity distribution is formed on the exit surface of the fly-eye integrator 6. As a result, as shown by the broken line portion 6b in FIG. 55, a secondary light source corresponding to the basic illumination area of a predetermined shape (for example, an annular shape) is formed at the incident surface position of the fly-eye integrator 6. It can be seen from the condenser lens system 82 side as if it were present. In other words, the secondary light source 6b as a virtual image
Are formed at the incident surface position of the fly-eye integrator 6. Therefore, it is sufficient to illuminate the reticle 9 via the condenser lens system 82 with the incident surface position of the fly-eye integrator 6 as the secondary light source formation surface. In consideration of laser resistance, it is preferable to divide the fly-eye integrator 6 and position the laser focusing point between the divided lens elements.

In each of the above-described embodiments, since the basic illumination area having the predetermined shape is formed on the entrance surface of the fly-eye integrator 6, it is composed of a large number of light sources discretely formed on the rear focal plane. The outline boundary line of the secondary light source having a predetermined shape is affected by the cross-sectional shape of the lens elements that form the fly-eye integrator 6, and an element that requires an aperture stop to limit the secondary light source to a desired shape. Becomes On the other hand, in the second modified example, since the basic illumination area having a predetermined shape is formed in the center of the fly-eye integrator 6, a secondary light source as a surface light source having a predetermined shape is formed as a virtual image on the incident surface. It

Therefore, in the second modification, the outline boundary line of the secondary light source is not affected by the cross-sectional shape of the lens elements forming the fly-eye integrator 6, and the aperture stop for limiting the secondary light source is It becomes unnecessary. Second above
The configuration shown in the modification can be applied to the first to sixth embodiments as necessary. For details of the configuration and operation according to the above-described second modification, for example, refer to the specification of Japanese Patent Application No. 2001-57790 and the drawings.

The diffractive optical element according to the present invention has a basic structure different from that of the illumination optical device shown in FIG.
It can also be applied to the illumination optical device disclosed in JP-A-01-85293. Specifically, the diffractive optical element 60 (61) in the illumination optical device shown in FIG. 1 of the publication.
As the above, the diffractive optical element according to the present invention can be applied. The configuration shown in the above-described second modification can be applied to the illumination optical device disclosed in Japanese Patent Laid-Open No. 2001-85293.

The diffractive optical element according to the present invention is E
It can also be applied to the illumination optical device disclosed in Japanese Patent Publication No. P1014196A2. Specifically, FIG. 4 of the publication.
The diffractive optical element 100 in the illumination optical device shown in FIG.
The diffractive optical element according to the present invention can be applied as 4,1004b and 1004c. Note that EP101
In the illumination optical device shown in FIG. 43 of Japanese Patent No. 4196A2, instead of the concave conical prism 1005B and the convex conical prism 1005C, or in the concave conical prism 100.
In addition to 5B and the convex conical prism 1005C, a concave V-groove prism and a convex V-groove prism may be arranged.

Here, the concave V-groove prism has two planar refracting surfaces arranged so as to have a ridge line in the direction orthogonal to the optical axis, and these two planar refracting surfaces direct the concave surface to the downstream side. Is formed in. On the other hand, a convex V-groove prism has a concave V
It has two planar refracting surfaces arranged so as to have a ridge in a direction parallel to the ridge of the groove prism, and these two planar refracting surfaces are formed so that the convex surface faces the upstream side (light source side). ing. At least one of these two V-groove prisms is configured to be movable along the optical axis,
The distance between the concave refracting surface of the concave V-groove prism and the convex refracting surface of the convex V-groove prism is variable.

Also, the conical prism pair 1005 and V
In the configuration in which the pair of groove prisms is arranged in the illumination optical path, another pair of V groove prisms may be provided. In this case, the ridgeline of the second pair of V-groove prisms is
It is orthogonal to the ridgeline of the groove prism pair. Also, such a conical prism pair 1005 and two V-groove prism pairs are arranged, or a conical prism pair 100
A configuration in which 5 is omitted is also possible. Note that EP10141
The configuration shown in the above-described second modification can be applied to the illumination optical device disclosed in Japanese Patent Publication No. 96A2.

In each of the embodiments of the present invention described above, a fly-eye integrator or a microlens array in which a plurality of optical elements are integrated is used as an optical integrator. However, the present invention is not limited to the wavefront division type optical integrator, and an internal reflection type optical integrator can be used. When the rod type integrator is arranged as the internal reflection type optical integrator, a substantial surface light source composed of virtual images of a plurality of light source images is formed in the vicinity of the incident end face thereof.

(Modification of Sixth Embodiment) Next, as a modification of the sixth embodiment, similar to the sixth embodiment, the structure of a diffractive optical element for annular illumination used for annular illumination and its manufacture. The method will be described. The present modification is an embodiment in which an even more uniform illuminance distribution can be obtained in the wafer conjugate plane and the illumination pupil plane than in the sixth embodiment. The modification will be described below, focusing on the difference from the sixth embodiment.

FIG. 56 (a) is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions in the modification of the sixth embodiment, and FIG. 56 (b) is a diagram showing a large number of divided regions. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined corresponding to a partial area | region densely. In the sixth embodiment, FIG. 13 in the first embodiment is used.
The 51 partial regions of FIG. 13A and the 51 partial optical elements of FIG.
The 216 partial regions shown in FIG. 41A and the 216 partial optical elements shown in FIG.
The basic elements shown in are determined. At this time, in the 216 partial regions finally formed by the subdivision, the shape of the equally divided partial regions is a rectangular shape having a long side substantially along the x direction.

In this modification, the procedure of randomizing the partial regions and the partial optical elements is the same as in the sixth embodiment,
The sixth embodiment is different from the sixth embodiment in that the shape of the equally divided partial areas among the partial areas finally formed by the subdivision is set to the square shape of the corresponding partial optical element. That is, the sizes in the y direction of the equally-divided partial areas and the non-equally-divided partial areas are the same as in the sixth embodiment.
It is uniquely determined by the equally-spaced division pitch qy in the y direction when setting the division projection basic area PIA1 (see FIGS. 13 and 14). However, since the size qx of the equally divided partial area in the x direction has a degree of freedom in selection,
In this modification, the size qx of the equally divided partial area in the x direction is y.
It is set equal to the size qy in the direction.

As described above, FIG. 56 corresponds to FIG. 41 in the sixth embodiment, but in FIG. 41A, the upper half divided projection basic area PIA1 is divided into 216 partial areas. On the other hand, in FIG. 56A, the upper half divided projection basic area PIA1 is further subdivided into 548 partial areas. At the time of subdivision, most of the divided projection basic area PIA1 is divided into square equally divided partial areas of qy = qx, and the remaining portions are set as non-equally divided partial areas. That is, in this modification, the divided projection basic area PIA1 is subdivided into 548 partial areas, and the equal divided partial areas are set in a square shape of qy = qx, which is different from the sixth embodiment. is doing.

Further, in FIG. 56 (b), with respect to the 51 partial optical elements densely arranged in FIG. 13 (b) of the first embodiment, the equal division partial regions are set in a square shape of qy = qx. It is further subdivided into 548 partial optical elements. Here, in FIG. 56B, FIG.
The rule of subdividing the 51 partial optical elements into the 548 partial optical elements is the rule of subdividing the 51 partial regions of FIG. 13A into the 548 partial regions in FIG. Totally. Therefore, for example, the partial area 1 in FIG. 13A is divided into 14 equally divided partial areas and one non-equally divided partial area in FIG.
The partial optical element 1 in 3 (b) is divided into 14 equally divided partial optical elements and one non-equally divided partial optical element in FIG. 56 (b).

FIG. 57 is a diagram showing a state of random arrangement in which indexes 1 to 548 are added to the 548 partial areas of FIG. 56 (a). In addition, FIG. 58 shows 5 in FIG.
It is a figure which shows the mode of a random arrangement by attaching indexes 1-548 to 48 partial optical elements. The procedure for randomly arraying the 548 partial regions can be performed in the same manner as in the sixth embodiment. FIGS. 56A and 57 show the result of the random array in each column of each partial region after subdivision. Specifically, in the random arrangement in each column, random replacement is performed in each column extending along the x direction. As described above, by performing the random replacement of the equally divided partial areas between the two columns, it is possible to obtain a better random array.

Further, in FIGS. 56B and 58, the random arrangement of the partial optical elements after subdivision in units of columns, the random arrangement in each row, and the random replacement of the equal division partial optical elements between the rows are shown. The results of are shown. Specifically, in a column-by-column random array, columns are randomly replaced in units of columns extending along the x direction. Further, in the random arrangement within each row, the partial optical elements are randomly exchanged within each row extending along the x direction. Furthermore, in the random replacement of the equally-divided partial optical elements between the rows, the evenly-divided partial optical elements are randomly replaced between the two rows extending along the x direction.

As described above, in the modified example of the sixth embodiment, the 548 partial areas of FIG. 56 (a), which are subdivided and randomly arranged, and the origin O of FIG. 56 (a) or Oz set near the origin O are set. Is overlapped with the FZP pattern centered at, and the FZP pattern is divided according to the 548 partial regions, so that 548 partial optical elements are obtained. The 548 partial optical elements thus obtained are densely rearranged according to the random arrangement shown in FIG. 56 (b) to form the divided basic optical element P (1).

Further, in this modification, as in the other embodiments, the case where a phase type binary Fresnel zone plate is used as a predetermined optical element (virtual lens) which is a basis for setting the pattern of the basic optical element is described. I am assuming. In this case, the phase type binary Fresnel zone plate is
The + 1st-order diffracted light and the -1st-order diffracted light are generated from each local point of the phase type binary Fresnel zone plate, and the + 1st-order light and the -1st-order light are respectively the Fourier transform surface of the relay lens system 4. It has a characteristic of illuminating a point-symmetrical position with respect to the origin (optical axis) on (the entrance surface of the fly-eye integrator 6 or the like). Therefore, the upper half (y ≧
0) The divided basic optical element P (1) obtained by patterning the divided projected basic area PIA1 according to the above-mentioned procedure.
Can be directly applied as the basic optical element. That is, when a light beam is incident on the divided basic optical element P (1), not only the semi-annular zone area of FIG. 56 (a) but also the semi-zonal zone area obtained by vertically inverting the area of FIG. 56 (a) with respect to the x-axis is illuminated. can do. As a result, by using only the divided basic optical element P (1), it is possible to form the entire ring-shaped basic illumination area IA1 originally assumed on the incident surface of the fly-eye integrator 6.

Further, although the patterning procedure is the same as that of the divided basic optical element P (1), the basic optical element P (j) is different from the divided basic optical element P (1) in the form of the random arrangement of internal partial optical elements. ) Can be generated. That is, the basic optical element P (j) is subjected to the random arrangement according to the j-th form different from the first form of the random arrangement used when the divided basic optical element P (1) is generated.
Can be generated. In this modification, as will be described later, the generation operation of the divided basic optical elements described above is performed a plurality of times while changing the form of the random array, and the plurality of basic optical elements P (j ) Is formed and the plurality of basic optical elements P (j)
To form a pattern of the diffractive optical element.

FIG. 59 is a diagram for explaining a method for improving the symmetry of the illuminance distribution in the x direction in the modification of the sixth embodiment. In FIG. 59, the pattern of the basic optical element P (j) _x is generated by subjecting the pattern of the basic optical element P (j) to inversion in the x direction, that is, left / right inversion. Then, the basic optical element P (j)
And P (j) _x are arranged adjacent to each other in the x direction to form a basic optical element assembly AP (j). In this way, by arranging two basic optical element patterns having a reverse relationship in the x direction so as to be adjacent to each other in the x direction, the symmetry of the illuminance distribution in the x direction can be secured.

The above design for the basic optical element assembly AP (j) is repeated 16 times from j = 1 to 16 to obtain basic optical element assemblies AP (1) to AP (16). That is, for j = 2 to 16, basic optical elements P (j) are respectively generated according to a random arrangement different from the random arrangement of the first mode of j = 1 shown in FIGS. The basic optical element aggregate AP (j) is generated by applying the one-setting rule shown in FIG. Hereinafter, the basic optical element aggregate AP (j) will be simply referred to as APj.

FIG. 60 is a diagram schematically showing a pattern formed on a mask used for manufacturing a diffractive optical element in the modification of the sixth embodiment by lithography. Referring to FIG. 60, in this modification, two block patterns AAP1 and AAP2 are formed on the mask. Here, the block pattern AAP1 is generated by arranging eight basic optical element assemblies AP1 to AP8 adjacent to each other along the x direction. The block pattern AAP2 includes eight basic optical element aggregates AP9 to A9.
It is generated by arranging P16 adjacently along the x direction. In this way, a mask similar to that shown in FIG. 46 can be manufactured based on the block patterns AAP1 and AAP2. Specifically, two block patterns AAP1 and AAP2 shown in FIG. 60 are provided in the center of the mask.
Are formed by, for example, EB drawing. Further, three alignment marks, a pair of cutting guide patterns, and a line and space pattern as a control rule pattern are formed around the mask.

FIG. 61 is a diagram showing a diffractive optical element formed on a glass substrate using a mask manufactured based on the block pattern of FIG. Referring to FIG. 61, two block patterns A of a mask (not shown) are shown.
The pattern of the diffractive optical element of the present modification is formed by integrating AP1 and AAP2 by projection exposure according to a step-and-repeat method. Specifically, two block patterns AAP
The transfer patterns of 1 and AAP2 are formed alternately so as not to be adjacent to each other in the vertical direction in the drawing.
As a result, a plurality of block patterns AAP1 and AAP2 are included within the effective diameter of the diffractive optical element, and the index j indicating different random forms is once included in each of the AAP1 and AAP2 once for each index. Each included. That is, each of AP (1) to AP (16) is included once. Therefore, the diffractive optical element finally manufactured is designed so that the same number of basic optical elements corresponding to the respective types j are included in the effective diameter.

FIG. 62 is a diagram showing a simulation result of a light intensity distribution when the diffractive optical element according to the modification of the sixth embodiment is applied to the exposure apparatus of FIG. That is, the simulation result shown in FIG. 62 is obtained by arranging the diffractive optical element of the present modification as the diffractive optical element 21 in the exposure apparatus of FIG. 7, shaping the excimer laser light with a wavelength of 248 nm into a rectangular section of 20 mm × 40 mm, and Illumination distribution (light intensity distribution) formed on the entrance surface of the fly-eye integrator 6 by the light flux generated by the diffractive optical element via the relay lens system 4 when entering the optical element.
Is. In addition, on the illumination pupil plane near the exit surface of the fly-eye integrator 6, an illumination distribution (light intensity distribution) that substantially corresponds to the illumination distribution (light intensity distribution) formed on the incident surface.
Is formed. In this modification, an illumination distribution (light intensity distribution) along each contour line of a large number of lens elements forming the fly-eye integrator 6 is formed on the incident surface of the fly-eye integrator 6. An aperture stop is arranged near the exit surface of the integrator 6 to define the shape of the illumination pupil.

Referring to FIG. 62, in the present modification, FIG.
A uniform light intensity distribution similar to the outer shape of the divided projection basic area PIA1 of 6 (a) can be formed on the entrance surface of the fly-eye integrator 6 and further on the illumination pupil surface.
In particular, in this modification, most of the partial optical elements are set to have a symmetrical shape in the x direction and the y direction, that is, a square shape. Almost uniform in the x direction and the y direction, and good imaging performance and optical performance can be secured.

That is, qx =
By setting qy, the anisotropy of the diffraction characteristics in the x direction and the y direction of the light flux from each partial optical element is reduced, and by extension, xy of the light intensity distribution on a predetermined surface such as the incident surface of the fly-eye integrator is reduced. Asymmetry is reduced,
Furthermore, the xy asymmetry of the pupil distribution is reduced. Further, as in a modified example of the seventh embodiment described later, a basic optical element based on a certain projection basic area and a basic optical element based on a projection basic area obtained by rotating the projection basic area by 90 degrees are used together. In the form, the light intensity distribution x
Good illumination with reduced y-asymmetry can be achieved.
Even when qx = qy is not satisfied, as shown in the eighth embodiment, a basic optical element based on a certain projection basic area and a basic optical element based on a projection basic area obtained by rotating the projection basic area by 90 degrees It is possible to obtain a predetermined xy asymmetry reduction effect by using in combination.

Further, as described above, since the phase type binary Fresnel zone plate is used as the predetermined optical element (virtual lens), the divided basic optical element P (1) corresponding to the upper half divided projection basic area PIA1. ~ P (1
It is actually possible to form the entire ring-shaped basic illumination area IA1 using only 6).

(Modified Example of Seventh Embodiment) Next, as a modified example of the seventh embodiment, the structure of a diffractive optical element for annular illumination used in annular illumination as in the seventh embodiment and a method of manufacturing the same. Will be described. The present modification is an embodiment in which an illuminance distribution having higher uniformity than the seventh embodiment can be obtained in the wafer conjugate plane and the illumination pupil plane. The modification will be described below, focusing on the differences from the seventh embodiment.

In this modified example, as in the seventh embodiment, a microlens array (not shown) is used instead of the fly-eye integrator 6, and an aperture stop is used at or near the illumination pupil plane. Instead, the objective is to generate a basic illumination area having a shape approximately equal to the desired pupil shape itself. However, as a diaphragm for the purpose of flare cutting, for example, a diaphragm having a slight margin outside the desired pupil shape may be used.

FIG. 63 (a) is a diagram showing a state in which the upper half divided projection basic area is divided into a large number of partial areas in the modification of the seventh embodiment, and FIG. 63 (b) is a drawing showing a large number of divided areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined corresponding to a partial area | region densely. Note that FIGS. 63A and 63B show a state in which the subdivision and the random arrangement have been completed. In addition, each partial region (or partial optical element) before densification included in FIG. 63A and each partial region (or partial optical element) after densification included in FIG. 63B are actually 57 and 58 of the modified example of the sixth embodiment, there is a one-to-one correspondence, but the reference numerals are omitted here to avoid complication. In this modified example, assuming the arrangement corresponding to FIG. 48 as in the seventh embodiment, a ring-shaped virtual transmission region IAP of the virtual ring aperture stop is assumed.
A basic illumination area IA301 (not shown) having a boundary line as close to a circle as possible along the shape of 301 (not shown) is set.

Specifically, the basic illumination area IA301 is
Half area (0 ≦ y) of the ring-shaped virtual transparent area IAP301
Is quantized in 22 steps along the y direction to approximate the shape to the shape of the ring-shaped virtual transmission region IAP301. In the present modification, a ring-shaped projection basic area PIA301 is obtained by projecting a ring-shaped basic illumination area IA301 obtained by 22-step quantization on a predetermined virtual diffraction lens. By the way, in the seventh embodiment, the basic illumination area IA300 is a ring-shaped virtual transmission area I.
The half-area (0 ≦ y) of AP is quantized in 43 steps along the y-direction, so that its shape is a ring-shaped virtual transmission area IA.
It is approximated to the shape of P. In addition, in the seventh embodiment, 2
A ring-shaped secondary light source having a ring ratio of / 3 (inner diameter / outer diameter) is formed on the illumination pupil surface, but in this modification, a ring-shaped secondary light source having a ring ratio of 1/2 is formed. Is formed.

Here, the subdivision of the partial region and the partial optical element is performed in the same manner as in the modification of the sixth embodiment described above, and the shape of the equally divided partial region among the partial regions finally generated by the subdivision. Is set to a square, and the shape of the equally divided partial optical element among the partial optical elements finally generated by the subdivision is set to a square. That is, FIG.
As shown in (a), the equally divided partial areas are set to have a square shape of qx = qy. As described above, by subdividing the equally divided partial areas while setting the restriction that the partial areas are set in a square shape, in the present modification, the number of partial areas and partial optical elements included in FIGS. 63A and 63B is included. Has finally reached 549.

The procedure for randomly arraying 549 partial regions and 549 partial optical elements is as follows.
It can be performed similarly to the sixth embodiment and its modification. That is, by performing random exchange within each column extending along the x direction (random array in each column), and further performing random exchange of evenly divided partial areas between two columns, if necessary,
The result of the random arrangement of the partial regions as shown in FIG. 63 (a) is obtained. In addition, by the random arrangement of the partial optical elements in units of columns, the random arrangement in each row, and the random replacement of the equally divided partial optical elements between the rows,
The result of the random arrangement of the partial optical elements as shown in (b) is obtained.

Thus, in this modification, the 549 partial areas of FIG. 63 (a), which are subdivided and randomly arranged, and the origin O of FIG. 63 (a) or O set near the origin O are set.
By overlapping the FZP pattern centered at z and dividing the FZP pattern according to the 549 partial regions, 549 partial optical elements are obtained.
The obtained 549 partial optical elements are shown in FIG.
The divided basic optical elements P (1) are densely rearranged according to the random arrangement shown in (b).

Further, in this modification, a predetermined optical element (virtual lens) from which the pattern of the basic optical element is set is set.
Assuming that a phase type binary Fresnel zone plate is used as. Therefore, also in this modification, as in the modification of the sixth embodiment, the upper half (0 ≦
y) The divided basic optical element P obtained by patterning only the divided projection basic area PIA301 in the above procedure.
It is possible to use (1) as it is as a basic optical element.

By the way, the basic optical element P (1) generated as shown in FIG. 63 (b) has a dimension L along the x direction.
It has a dimension Lsp_y along the sp_x and y directions. The x-direction dimension Lsp_x and the y-direction dimension Lsp_y are designed to be substantially equal, but the y-direction dimension Lsp_y is
Is slightly larger than the dimension Lsp_x in the x direction,
As a result, the basic optical element P (1) has a rectangular shape with long sides along the y direction. The basic optical element P
(1) x-direction dimension Lsp_x and y-direction dimension Lsp
_Y is a predetermined value that depends on the value of the focal length funi of the virtual lens and the value of the focal length fL of the relay lens system 4 shown in FIG.

According to the principle of the present invention, the ratio f of fL to funi
By changing L / funi, the basic optical element P
(1) x-direction dimension Lsp_x and y-direction dimension Lsp
It can be easily inferred that _y is proportionally reduced or proportionally enlarged. Therefore, in this modification, this property is used to
The patterning of the basic optical element assembly is advanced so that the diffraction characteristics are substantially symmetrical with respect to the y-direction and the y-direction. That is, in FIG. 63 (a), since the original quantization of the partial area and further the quantization of the partial optical element are performed in the y direction, and then the subdivision and the random arrangement are performed, the boundary of the divided projection basic area PIA301 is The line is a region where the absolute value of the y coordinate is large, and the circular approximation is slightly worse.

In order to improve the circular approximation of the boundary line of the projection basic area which defines this ring-shaped basic illumination area, in this modification, the basic optical element designed by quantization along the x direction and the y direction are used. A method is adopted in which the same number of basic optical elements designed by quantization along the same direction are mixed within the effective diameter of the diffractive optical element. As a result, an illumination without quantization anisotropy, that is, a ring-shaped basic illumination region having a good circular approximation can be formed on the illumination pupil plane. In order to carry out this method, first, the same division and random arrangement as in the above-mentioned divided basic optical element P (1) are performed, but a virtual lens when the divided basic optical element P (1) is patterned. Of the divided basic optical element P which is patterned assuming a virtual lens having a focal length funi 'different from the focal length funi of
Consider (1) '. The divided basic optical element P (1) '
Focal length f of the relay lens system 4 when patterning
L has the same value as in the case of the divided basic optical element P (1).

FIG. 64A shows a state in which the upper half divided projection basic region is divided into a large number of partial regions assuming a virtual lens having a focal length different from that in FIG. 63 in the modification of the seventh embodiment. 64B is a diagram showing a divided basic optical element obtained by densely rearranging a large number of partial optical elements defined corresponding to a large number of divided partial areas. That is, FIGS. 64 (a) and 64 (b) depend on the magnification factor (funi / funi ′), and FIG.
And a divided projection basic area PIA301 ′ similar to (b)
And a divided basic optical element P (1) '.

The patterning in the divided basic optical element P (1) 'is performed by projecting the basic illumination area IA301 onto a predetermined virtual diffraction lens according to the principle described with reference to FIG. In order to obtain, a virtual lens having a focal length funi 'and a relay lens system 4 having a focal length fL are used. This can be done by applying exactly the same patterning history as in 1).

The x-direction dimension Lsp_x ′ and the y-direction dimension Lsp_y ′ of the divided basic optical element P (1) ′ thus patterned are the x-direction dimension of the divided basic optical element P (1) according to the above-described magnification principle. It is expressed by the following equations (9) and (10) using Lsp_x and the y-direction dimension Lsp_y. Incidentally, the x-direction dimension qx ′ and the y-direction dimension qy ′ of the equally-divided partial optical element in the divided basic optical element P (1) ′ are also equal to the equally-divided partial optical elements in the divided basic optical element P (1) according to the above-described magnification principle. It is expressed by the following equations (9 ') and (10') using the element's x-direction dimension qx and y-direction dimension qy.

[0251]   Lsp_x '= (funi / funi') Lsp_x (9)   Lsp_y '= (funi / funi') Lsp_y (10)   qx '= (funi / funi') qx (9 ')   qy '= (funi / funi') qy (10 ')

FIG. 65A shows a state in which the divided projection basic region obtained by rotating the divided projection basic region of FIG. 64 by 90 degrees in the xy plane in the modification of the seventh embodiment is divided into a number of partial regions. FIG. 65 (b) shows a divided basic optical element obtained by densely rearranging a large number of partial optical elements defined corresponding to a large number of partial areas divided in (a). It is a figure. In this modification, as shown in FIG.
The divided projection basic area P (1) ′ in FIG. 64 is set to 90 in the xy plane.
Degree basic optical element P (1) '_ 90 obtained by rotating
Assume Here, the divided basic optical element P (1) ′ _ 9
The dimension of 0 in the x direction is Lsp_y ′, and the dimension in the y direction is Lsp_y ′.
sp_x '. The divided basic optical element P (1) '
In Lsp_x ′ and Lsp_y ′ indicating the dimension of _90, the dimension notation of the split basic optical element P (1) ′ before being rotated by 90 degrees is used as it is in order to prevent unnecessary confusion.

Then, in order to arrange the divided basic optical element P (1) and the divided basic optical element P (1) '_ 90 adjacent to each other without a gap along the x direction, the divided basic optical element P (1) in the y direction. Dimension Lsp_y and divided basic optical element P (1) '_
It is set to satisfy the following relational expression (11) so that the y-direction dimension Lsp_x ′ of 90 matches. Lsp_
x '= Lsp_y (11)

Thus, the relationship shown in the following expression (12) is obtained based on the expressions (9) to (11). Then, from the equation (12), the required focal length funi ′ of the virtual lens for satisfying the relational equation (11) is represented by the equation (13).
In Expression (13), the right sides are all known quantities, and the required focal length funi ′ of the virtual lens is uniquely determined. (Funi / funi ') Lsp_x = Lsp_y (12) funi' = funi · Lsp_x / Lsp_y (13)

Thus, the required focal length fu of the virtual lens is
By defining ni ′, the divided basic optical element P (1)
Y-direction dimension Lsp_y and the divided basic optical element P
(1) ′ can be matched equally with the y-direction dimension Lsp_x ′ of the divided basic optical element P (1) ′ _ 90 obtained by rotating 90 ° in the xy plane, resulting in two divided basic optical elements. It is possible to arrange P (1) and P (1) ′ — 90 adjacent to each other in the x direction, and thus in a two-dimensional and dense manner.

Referring to FIG. 65 (a), FIG. 64 (a)
Area with a large absolute value of the y coordinate in the divided projection basic area PIA301 ′ (that is, an area where the circular approximation is slightly bad)
, Is rotationally moved to a region in which the absolute value of the x coordinate is large in the divided projection basic region PIA ′ — 90. That is, the divided basic optical element P (1) ′ _ 90 can be considered as a basic optical element designed by quantization along the x direction. Therefore, the split elementary optical element P (1) ′ _ 90 designed by the quantization along the x direction and the original split elementary optical element P designed by the quantization along the y direction.
By mixing (1) and (1) in substantially the same number within the effective diameter of the diffractive optical element, the illumination in which the anisotropy of quantization is significantly reduced, that is, the ring-shaped basic illumination area having a significantly good circular approximation is illuminated. Can be formed on the pupil plane.

Furthermore, the divided basic optical element P (1) of FIG. 63 (b) and the divided basic optical element P of FIG. 65 (b).
The patterning procedure is the same as that of (1) ′ _ 90, but the basic optical element P (j) is different from the divided basic optical elements P (1) and P (1) ′ _ 90 in the form of the random arrangement of the partial optical elements inside. And P (j) '_ 90 can be generated. That is, the basic optical elements are arranged through the random arrangement according to the j-th form different from the first form of the random arrangement used to generate the divided basic optical elements P (1) and P (1) ′ _ 90. P (j) and P
(J) '_ 90 can be generated. In this modification, as will be described later, the generation operation of the divided basic optical elements described above is performed a plurality of times while changing the form of the random array, and the plurality of basic optical elements P (j ) And P (j) ′ _ 90 and form a plurality of these basic optical elements P (j) and P (j) ′ _ 90.
(J) '_ 90 is integrated to form a pattern of the diffractive optical element.

FIG. 66 illustrates a method of forming the basic optical element aggregate AP (j) by arranging the basic optical elements P (j) and P (j) '_ 90 adjacent to each other in the modification of the seventh embodiment. FIG. As described above, the divided basic optical element P (j) is a rectangle having a long side in the y direction, but in the divided basic optical element P (j) ′ _ 90, the divided basic optical element P (j) is in the xy plane. It is rotated 90 degrees and is proportionally expanded to have a dimension Lsp_x ′ in the y direction and a split basic optical element P.
Since the y-direction dimension Lsp_y of (j) is matched, the divided basic optical elements P (j) and P (j) ′ _ 90 can be arranged adjacent to each other along the x direction without any gap. Incidentally, in FIG. 66, in order to facilitate understanding of the explanation, the divided basic optical element P which is actually a rectangle close to a square is used.
(J) and P (j) ′ — 90 are exaggeratedly displayed in a rectangular shape that is elongated to some extent.

Thus, the basic optical elements P (j) and P (j)
The basic optical element aggregate AP (j) is formed as an aggregate obtained by arranging (j) ′ _ 90 and the (j) ′ _ 90 adjacent to each other in the x direction. Then, the above-described design regarding the basic optical element assembly AP (j) is repeated 12 times from j = 1 to 12 to obtain basic optical element assemblies AP (1) to AP (12). That is, for j = 2 to 12, basic optical elements P (j) are generated according to a random arrangement different from the random arrangement shown in FIG. 63, and basic optical elements P (j) ′ _ 90 are generated for each of them. And each j
The basic optical element aggregate AP (j) is generated by applying the one-setting rule shown in FIG. Hereinafter, the basic optical element aggregate AP (j) will be referred to as APj.

FIG. 67 illustrates a method of forming the basic optical element assembly BP (j) by arranging the basic optical elements P (j) and P (j) '_ 90 adjacent to each other in the modification of the seventh embodiment. FIG. 66 and 67, the basic optical element assembly AP (j) is arranged by arranging the basic optical element P (j) ′ _ 90 adjacently on the right side of the basic optical element P (j) in FIG. 66. While forming
In FIG. 67, the basic optical element assembly BP (j) is formed by arranging the basic optical element P (j) adjacently on the right side of the basic optical element P (j) ′ _ 90 in the drawing. In other words,
In the basic optical element aggregates AP (j) and BP (j), the arrangement order of the basic optical elements P (j) and P (j) ′ _ 90 along the x direction is opposite.

Then, the above design for the basic optical element assembly BP (j) is repeated 12 times from j = 13 to 24 to obtain the basic optical element assembly BP (13) to BP (24). That is, for j = 13 to 24, j =
The basic optical elements P (j) are respectively generated according to a random arrangement different from the random arrangements of 1 to 12, basic optical elements P (j) ′ _ 90 are generated for each of them, and one for each j is shown in FIG. The basic optical element aggregate BP (j) is generated by applying the setting rule. Hereinafter, the basic optical element assembly BP (j) will be referred to as BPj.

FIG. 68 is a diagram partially showing the pattern of the basic optical element P (1) in the modification of the seventh embodiment. Further, FIG. 69 is a diagram partially showing the pattern of the basic optical element P (2) in the modification of the seventh embodiment. Specifically, in FIG. 63B, only the boundary lines of the partial optical elements are shown for simplicity, but FIG.
About about 6 × 6 partial optical elements in the lower left corner of the basic optical element P (1) shown in (b) (that is, basic optical elements in which partial optical elements are randomly arranged according to the first mode) The details of the pattern are also displayed.

Further, FIG. 69 shows about 6 × 6 pieces at the lower left corner of the basic optical element P (2) (that is, basic optical elements in which partial optical elements are randomly arranged according to the second mode) which is not shown. The partial optical element of is also shown including the details of its internal pattern. In FIG. 68 as well, the boundary of the partial optical elements can be recognized,
Observing along this boundary, it can be seen from the two figures that the boundary of the partial optical element in FIG. 68 and the boundary of the partial optical element in the lower left corner of FIG. 63 (b) have a one-to-one correspondence in shape and arrangement. It can be confirmed by comparison. In other words, correct correspondence is made between the two figures, including the positional relationship between the equally divided partial optical element and the non-uniformly divided partial optical element.

The internal pattern of the partial optical element shown in FIGS. 68 and 69, or the internal pattern of the partial optical element similar to this, is the same as the design procedure in all the other embodiments including the first embodiment. A phase type binary Fresnel zone plate or the like is used as a virtual lens, a required portion is set as in each embodiment, a partial optical element is set according to the required portion, and the virtual lens is divided and re-aligned along the boundary of the partial optical element. It is the result of arranging, subdividing and randomizing as needed. Therefore, FIG.
8 and the pattern contained in each of the partial optical elements in FIG. 69 is a pattern obtained by cutting out a part of a phase type binary Fresnel zone plate or a ring-shaped pattern having a similar action and arranging it at a predetermined position. Can be seen from the figure.

68 and 69 show the results when a binary phase type Fresnel zone plate is applied as a virtual lens. For example, in FIG. 68, the white portion of the pattern is the binary phase type Fresnel zone plate. The convex area portion and the black portion indicate the concave area portion, respectively. Here, the term binary phase type Fresnel zone plate is
This is the same as what is simply referred to as the "phase type binary Fresnel zone plate" in the above and the following description, but this point and related matters will be described in detail later.

Therefore, the black portions in FIGS. 68 and 69 do not block light, but merely display the concave area portion. Both the convex area portion and the concave area portion are made of a transparent material such as synthetic quartz glass. Comparing FIG. 68 and FIG. 69, it can be seen that the arrangement of the partial optical elements and the contents of the zone plate pattern inside each partial optical element are different, although they show almost the same area of the basic optical element. . This means that the basic optical element P (1) of FIG. 68 and the basic optical element P (2) of FIG. 69 have different random arrangements.

The basic optical element P (j) in which the partial optical elements are randomly arranged according to the j-th form can also display the same drawings as FIGS. 68 and 69, but the partial optical elements concerning the different form of j are shown. And the contents of the zone plate pattern inside the partial optical element are randomly different as described in the present invention.
These basic optical elements having different random arrangement forms diffractively different light intensity distributions on a predetermined surface (for example, the entrance surface of the fly-eye integrator 6), and thus the different basic optical elements are diffracted. By including a plurality of optical elements within the effective diameter, it is possible to perform illumination in which interference noise is favorably reduced.

FIG. 70 is a diagram schematically showing a pattern formed on a mask used for manufacturing a diffractive optical element in the modification of the seventh embodiment by lithography. Referring to FIG. 70, in the present modification, one block pattern AAP is formed on the mask. Here, the block pattern AAP includes twelve basic optical element assemblies AP1 to AP12 and twelve basic optical element assemblies BP1.
3 to BP24 are arranged adjacent to each other. More specifically, in the block pattern AAP, in order from the top, subblocks AP1 to AP6 are adjacently arranged along the x direction, subblocks BP13 to BP18 are adjacently arranged along the x direction, and AP7 to AP12 are x. Sub-blocks adjacent along the direction, and B
Subblocks in which P19 to BP24 are adjacently arranged along the x direction are adjacently arranged along the y direction.

In this way, a mask similar to that shown in FIG. 46 can be manufactured based on the block pattern AAP.
Specifically, one block pattern AAP shown in FIG. 70 is formed in the center of the mask by, for example, EB drawing. Further, three alignment marks, a pair of cutting guide patterns, and a line and space pattern as a control rule pattern are formed around the mask.

FIG. 71 is a diagram showing a diffractive optical element formed on a glass substrate using a mask manufactured based on the block pattern of FIG. Referring to FIG. 71, one block pattern A of a mask (not shown)
The pattern of the diffractive optical element of this modification is formed by integrating the AP by projection exposure according to the step-and-repeat method. Specifically, the transfer pattern by one block pattern AAP is formed vertically and horizontally. As a result, a plurality of block patterns AAP are included within the effective diameter of the diffractive optical element, but the index j indicating different random forms is included once for each of 1 to 24 in this AAP pattern. . That is, AP (1) to AP (12) and BP
(13) to BP (24) are included once each. Therefore, within the effective diameter of the finally manufactured diffractive optical element,
It is designed so that the same number of basic optical elements corresponding to the respective types of j are included.

FIG. 72 is a diagram showing a simulation result of a light intensity distribution when the diffractive optical element according to the modification of the seventh embodiment is applied to the exposure apparatus of FIG. That is, the simulation result shown in FIG. 72 is obtained by arranging the diffractive optical element of this modification as the diffractive optical element 21 in the exposure apparatus of FIG. 7, shaping the excimer laser light having a wavelength of 248 nm into a rectangular section of 20 mm × 40 mm, and then diffracting it. Illumination distribution (light intensity distribution) formed on the entrance surface of the fly-eye integrator 6 by the light flux generated by the diffractive optical element via the relay lens system 4 when entering the optical element.
Is. In addition, on the illumination pupil plane near the exit surface of the fly-eye integrator 6, an illumination distribution (light intensity distribution) that substantially corresponds to the illumination distribution (light intensity distribution) formed on the incident surface.
Is formed.

Referring to FIG. 72, in the present modification, FIG.
3 (a) divided projection basic area PIA301 and FIG.
It can be seen that it is possible to form a uniform light intensity distribution on the illumination pupil plane that is similar to the composite outline of the divided projection basic area PIA301′_90 in (a). In particular, in this modification, the basic optical element P (j) designed by quantization along the y direction and the basic optical element P (j) ′ _ 90 designed by quantization along the x direction are used as diffractive optical elements. Since a method of mixing substantially the same number within the effective diameter of 1 is applied, it is possible to form illumination without quantization anisotropy, that is, a ring-shaped basic illumination area having a good circular approximation on the illumination pupil plane.

In this modification, the basic optical element P
Two basic optical element aggregates AP (j) and BP in which (j) and P (j) ′ _ 90 are arranged in the opposite order in the x direction.
Since (j) is used, a more uniform light intensity distribution can be realized by the averaging effect. Further, in this modification, most of the partial optical elements are set to have a symmetrical shape in the x-direction and the y-direction, that is, a square shape, so that the sharpness of the edge of the illuminance distribution obtained on the wafer conjugate plane and the illumination pupil plane is small. Almost uniform in the x direction and the y direction, and good imaging performance and optical performance can be secured.

Further, as described above, since the phase type binary Fresnel zone plate is used as the predetermined optical element (virtual lens), the divided basic optical element AP (1) corresponding to the upper divided projection basic area PIA301. ~
AP (12) and right half (0 ≦ x) divided projection basic area P
Split basic optical element BP corresponding to IA301′_90
It is actually possible to form the entire ring-shaped basic illumination area IA1 by using only (13) to BP (24).

(Ninth Embodiment) Next, as the ninth embodiment, the structure of a diffractive optical element used for illumination in a diaphragmless illumination system and the method for manufacturing the same will be described as in the seventh embodiment and its modification. To do. However, in the ninth embodiment, unlike the seventh embodiment and its modification, the present invention is applied to a diffractive optical element for circular illumination used in circular illumination. The ninth embodiment will be described below, focusing on the differences from the modification of the seventh embodiment.

FIG. 73 (a) is a diagram showing a state in which the upper half divided projection basic area is divided into a large number of partial areas in the ninth embodiment, and FIG. 73 (b) shows a plurality of divided partial areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.
Note that FIGS. 73A and 73B show a state in which the subdivision and the random arrangement have been completed. In addition, FIG.
Each partial area (or partial optical element) before densification included in (a) and each partial area (or partial optical element) after densification included in FIG. As shown by the symbols in FIGS. 57 and 58 of the modified example of the sixth embodiment,
There is a one-to-one correspondence, but the symbols are not shown here to avoid complication. In the ninth embodiment,
A basic illumination area (not shown) having a boundary line as close to a circle as possible is set along the shape of the circular virtual transmission area of the virtual circular aperture stop.

Specifically, the basic illumination area is quantized along the y direction in the half area (0 ≦ y) of the circular virtual transparent area, so that the shape of the virtual virtual transparent area is changed. Is approximated to. In the present embodiment, by projecting a circular basic illumination area obtained by quantizing along the y direction onto a predetermined virtual diffractive lens, the circular projection basic area can be converted into the projection basic area shown in FIG. A semi-circular divided projection basic area as shown in FIG. The subdivision of the partial regions and the partial optical elements in the ninth embodiment is performed in the same manner as in the modification of the seventh embodiment described above, and the shape of the equally divided partial regions among the partial regions finally generated by the subdivision is set. It is set to a square, and the shape of the equally divided partial optical element among the partial optical elements finally generated by the subdivision is set to a square.

The procedure for randomly arranging the partial regions and the partial optical elements can be the same as in the seventh embodiment and its modification. That is, by performing random exchange within each column extending along the x direction, and further performing random exchange of the equally-divided partial regions between two columns, if necessary,
The result of the random array of partial regions as shown in FIG. 73 (a) is obtained. In addition, by the random arrangement of the partial optical elements on a column-by-row basis, the random arrangement within each row, and the random replacement of the equally divided partial optical elements between the rows, FIG.
The result of the random arrangement of the partial optical elements as shown in (b) is obtained.

Thus, in the ninth embodiment, the subregions of FIG. 73 (a) that are subdivided and randomly arranged and the partial regions of FIG.
The FZP pattern centered on the origin O or the Oz set near the origin O in (a) is overlapped to
By dividing the P pattern according to each partial area,
Each partial optical element is obtained. Further, the obtained partial optical elements are densely rearranged according to the random arrangement shown in FIG. 73 (b), and the divided basic optical element P (1) is formed.

Further, in the ninth embodiment, it is assumed that the phase type binary Fresnel zone plate is used as a predetermined optical element (virtual lens) which is a basis for setting the pattern of the basic optical element. Therefore, also in the present embodiment, similar to the modification of the seventh embodiment, the divided basic optical element P (1 obtained by patterning only the upper half (0 ≦ y) divided projected basic regions by the above-described procedure. ) Can be used as it is as a basic optical element.

Further, although the patterning procedure is the same as that of the divided basic optical element P (1), the basic optical element P (j) is different from the divided basic optical element P (1) in the form of the random arrangement of the internal partial optical elements. ) Can be generated. That is, the basic optical element P (j) is subjected to the random arrangement according to the j-th form different from the first form of the random arrangement used when the divided basic optical element P (1) is generated.
Can be generated. In the present embodiment, the generation operation of the divided basic optical element described above is performed 30 times while changing the form of the random array, and 30 basic optical elements P (1) to P (P) to P (1) to P
(30) are formed, and these 30 basic optical elements P are formed.
A pattern of the diffractive optical element is formed by integrating (1) to P (30) so that the same number is included in the effective diameter of the diffractive optical element for each type j.

FIG. 74 is a diagram showing a simulation result of the light intensity distribution when the diffractive optical element of the ninth embodiment is applied to the exposure apparatus of FIG. That is, the simulation result shown in FIG. 74 is obtained by arranging the diffractive optical element of the present embodiment as the diffractive optical element 21 in the exposure apparatus of FIG.
The illumination distribution (light intensity distribution) formed by the diffractive optical element on the incident surface of the fly-eye integrator 6 via the relay lens system 4 when the diffractive optical element is incident on the diffractive optical element after being shaped into a rectangular section of 0 mm. ). In addition, on the illumination pupil plane near the exit surface of the fly-eye integrator 6, the illumination distribution (light intensity distribution) formed on the entrance surface
An illumination distribution (light intensity distribution) that substantially corresponds to is formed.

Referring to FIG. 74, in the ninth embodiment,
It is understood that it is possible to form a uniform light intensity distribution on the illumination pupil plane, which is similar to the outer shape of the divided projection basic region in FIG. 73 (a). In particular, in the present embodiment, most of the partial optical elements are set to have a symmetrical shape in the x direction and the y direction, that is, a square shape, so that the sharpness of the edge of the illuminance distribution obtained on the wafer conjugate plane and the illumination pupil plane is small. x
It becomes almost uniform in the direction and the y direction, and good imaging performance and optical performance can be secured. Further, as described above, since the phase type binary Fresnel zone plate is used as the predetermined optical element (virtual lens), the divided basic optical element P corresponding to the upper half divided projection basic area is formed.
It is actually possible to form the entire circular basic illumination area by using only (1) to P (30).

In the ninth embodiment, as in the modification of the seventh embodiment described above, in addition to the split basic optical element P (j), split basic optics obtained by rotating 90 degrees in the xy plane. It is preferable to introduce the element P (j) '_ 90. In this case, the basic optical element P (j) designed by the quantization along the y direction and the basic optical element P (j) ′ _ 90 designed by the quantization along the x direction are defined as the effective diameter of the diffractive optical element. By mixing substantially the same number in the inside, it is possible to form illumination without quantization anisotropy, that is, a circular basic illumination area having a good circular approximation in the illumination pupil plane.

(Tenth Embodiment) Next, as a tenth embodiment, the configuration of a diffractive optical element used for illumination in a diaphragmless illumination system and a method of manufacturing the same will be described as in the seventh embodiment and its modification. To do. However,
In the tenth embodiment, unlike the seventh embodiment and its modification, the present invention is applied to a diffractive optical element for quadrupole illumination used in quadrupole illumination. The tenth embodiment will be described below, focusing on the differences from the modification of the seventh embodiment.

FIG. 75 (a) is a diagram showing a state in which the upper half divided projection basic area in the tenth embodiment is divided into a large number of partial areas, and FIG. 75 (b) shows a plurality of divided partial areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely. Note that FIGS. 75 (a) and 75 (b) show a state in which subdivision and random arraying have been completed. In addition, each partial area (or partial optical element) before densification included in FIG. 75A and each partial area (or partial optical element) after densification included in FIG. The sixth
As shown by the reference numerals in FIGS. 57 and 58 of the modified example of the embodiment, there is a one-to-one correspondence, but the reference numerals are omitted here to avoid complication. In the tenth embodiment, a quadrupole basic illumination region (not shown) having a boundary line as close to a circle as possible along the shape of the quadrupole virtual transmission region of the virtual quadrupole aperture stop having four circular openings. ) Is set.

Specifically, the shape of the basic illumination area is approximated to the shape of the quadrupole virtual transmission area by quantizing the half area of the quadrupole virtual transmission area along the y direction. ing. In the present embodiment, the quadrupole basic illumination area obtained by quantizing along the y direction is projected onto a predetermined virtual diffractive lens, so that the quadrupole projection basic area, and eventually FIG. The upper half (0 <
The dipole-shaped divided projection basic region of y) is obtained. The subdivision of the partial regions and the partial optical elements in the tenth embodiment is performed in the same manner as in the modification of the seventh embodiment described above, and the shape of the equally divided partial regions among the partial regions finally generated by the subdivision is set. It is set to a square, and the shape of the equally divided partial optical element among the partial optical elements finally generated by the subdivision is set to a square.

The procedure for randomly arranging the partial regions and the partial optical elements can be the same as in the seventh embodiment and its modification. That is, by performing random exchange within each column extending along the x direction, and further performing random exchange of the equally-divided partial regions between two columns, if necessary,
The result of the random arrangement of the partial regions as shown in FIG. 75 (a) is obtained. In addition, by performing a random arrangement of the partial optical elements on a column-by-row basis, a random arrangement within each row, and a random replacement of the equally divided partial optical elements between the rows, FIG.
The result of the random arrangement of the partial optical elements as shown in (b) is obtained.

As described above, in the tenth embodiment, the subregions of FIG.
5 (a) is overlapped with the FZP pattern centered on the origin O or Oz set near the origin O, and F
Each partial optical element is obtained by dividing the ZP pattern according to each partial region. Further, the obtained partial optical elements are densely rearranged according to the random arrangement shown in FIG. 75 (b) to form the divided basic optical element P (1).

Further, in the tenth embodiment, it is assumed that the phase type binary Fresnel zone plate is used as a predetermined optical element (virtual lens) which is a basis for setting the pattern of the basic optical element. Therefore, also in the present embodiment, similar to the modification of the seventh embodiment, the divided basic optical element P (1 obtained by patterning only the upper half (0 ≦ y) divided projected basic regions by the above-described procedure. ) Can be used as it is as a basic optical element.

Further, although the patterning procedure is the same as that of the divided basic optical element P (1), the basic optical element P (j) is different from the divided basic optical element P (1) in the form of the random arrangement of internal partial optical elements. ) Can be generated. That is, the basic optical element P (j) is subjected to the random arrangement according to the j-th form different from the first form of the random arrangement used when the divided basic optical element P (1) is generated.
Can be generated. In the present embodiment, the generation operation of the divided basic optical element described above is performed 30 times while changing the form of the random array, and 30 basic optical elements P (1) to P (P) to P (1) to P
(30) are formed, and these 30 basic optical elements P are formed.
A pattern of the diffractive optical element is formed by integrating (1) to P (30) so that the same number is included in the effective diameter of the diffractive optical element for each type j.

FIG. 76 is a diagram showing a simulation result of a light intensity distribution when the diffractive optical element of the tenth embodiment is applied to the exposure apparatus of FIG. That is, the simulation result shown in FIG. 76 is obtained by arranging the diffractive optical element of this embodiment as the diffractive optical element 21 in the exposure apparatus of FIG.
The illumination distribution (light intensity distribution) formed by the diffractive optical element on the incident surface of the fly-eye integrator 6 via the relay lens system 4 when the diffractive optical element is incident on the diffractive optical element after being shaped into a rectangular section of 0 mm. ). In addition, on the illumination pupil plane near the exit surface of the fly-eye integrator 6, the illumination distribution (light intensity distribution) formed on the entrance surface
An illumination distribution (light intensity distribution) that substantially corresponds to is formed.

Referring to FIG. 76, in the tenth embodiment, it is possible to form a uniform light intensity distribution on the illumination pupil plane, which is similar to the outer shape of the divided projection basic region of FIG. 75 (a). Recognize. In particular, in the present embodiment, most of the partial optical elements are set to have a symmetrical shape in the x direction and the y direction, that is, a square shape, so that the sharpness of the edge of the illuminance distribution obtained on the wafer conjugate plane and the illumination pupil plane is small. Almost uniform in the x direction and the y direction, and good imaging performance and optical performance can be secured. Further, as described above, since the phase type binary Fresnel zone plate is used as the predetermined optical element (virtual lens),
It is actually possible to form the entire quadrupole basic illumination area by using only the divided basic optical elements P (1) to P (30) corresponding to the upper half divided projection basic areas. .

In the tenth embodiment, as in the modification of the seventh embodiment described above, in addition to the divided basic optical element P (j), the divided basic optics obtained by rotating 90 degrees in the xy plane. It is preferable to introduce the element P (j) '_ 90. In this case, the basic optical element P (j) designed by the quantization along the y direction and the basic optical element P (j) ′ _ 90 designed by the quantization along the x direction are defined as the effective diameter of the diffractive optical element. By mixing substantially the same number in the inside, it is possible to form illumination without quantization anisotropy, that is, a quadrupole basic illumination area having a good circular approximation in the illumination pupil plane.

(Eleventh Embodiment) Next, as the eleventh embodiment, the structure of a diffractive optical element used for illumination in a diaphragmless illumination system and the method for manufacturing the same will be described as in the seventh embodiment and its modification. To do. However,
In the eleventh embodiment, unlike the seventh embodiment and its modification, the present invention is applied to a diffractive optical element for annular illumination used for annular illumination having an annular ratio (inner diameter / outer diameter) close to 1. Applied. The eleventh embodiment will be described below, focusing on the differences from the modification of the seventh embodiment.

FIG. 77 (a) is a diagram showing a state in which the upper half divided projection basic area is divided into a large number of partial areas in the eleventh embodiment, and FIG. 77 (b) is a diagram showing a large number of divided partial areas. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely. 77 (a) and 77 (b) show a state in which subdivision and random arraying have been completed. In addition, each partial region (or partial optical element) before densification included in FIG. 77A and each partial region (or partial optical element) after densification included in FIG. 77B are actually The sixth
As in the relationship between FIG. 57 and FIG. 58 of the modified example of the embodiment,
There is a one-to-one correspondence with the reference numerals, but the reference numerals are omitted here for the sake of simplicity. 11th
In the embodiment, a basic illumination region (not shown) having a boundary line that is as close to a circle as possible is set along the shape of the ring-shaped virtual transmission region of the virtual ring aperture stop.

Specifically, the basic illumination area is quantized along the y direction in a half area (0 ≦ y) of the ring-shaped virtual transmission area, so that its shape is the shape of the ring-shaped virtual transmission area. Is approximated to. In the present embodiment, by projecting a ring-shaped basic illumination area obtained by quantizing along the y direction onto a predetermined virtual diffractive lens, the ring-shaped projected basic area, and eventually FIG. A semi-ring shaped divided projection basic area as shown in FIG. The subdivision of the partial regions and the partial optical elements in the eleventh embodiment is performed in the same manner as in the modification of the seventh embodiment described above, and the shape of the equally divided partial regions among the partial regions finally generated by the subdivision is set. It is set to a square, and the shape of the equally divided partial optical element among the partial optical elements finally generated by the subdivision is set to a square.

The procedure for randomly arranging the partial regions and the partial optical elements can be the same as in the seventh embodiment and its modification. That is, by performing random exchange within each column extending along the x direction, and further performing random exchange of the equally-divided partial regions between two columns, if necessary,
The result of the random arrangement of the partial regions as shown in FIG. 77 (a) is obtained. 77. Further, by performing a random arrangement of the partial optical elements in units of columns, a random arrangement in each row, and a random replacement of the equally divided partial optical elements between the rows, FIG.
The result of the random arrangement of the partial optical elements as shown in (b) is obtained.

Thus, in the eleventh embodiment, the subregions of FIG. 77 (a) that are subdivided and randomly arranged are compared with the partial regions of FIG.
7 (a) is overlapped with the FZP pattern centered on the origin O or Oz set near the origin O, and F
Each partial optical element is obtained by dividing the ZP pattern according to each partial region. Further, the obtained partial optical elements are densely rearranged according to the random arrangement shown in FIG. 77 (b), and the divided basic optical element P (1) is formed.

Further, in the eleventh embodiment, it is assumed that the phase type binary Fresnel zone plate is used as a predetermined optical element (virtual lens) which is a basis for setting the pattern of the basic optical element. Therefore, also in the present embodiment, similar to the modification of the seventh embodiment, the divided basic optical element P (1 obtained by patterning only the upper half (0 ≦ y) divided projected basic regions by the above-described procedure. ) Can be used as it is as a basic optical element.

Further, although the patterning procedure is the same as that of the divided basic optical element P (1), the basic optical element P (j) is different from the divided basic optical element P (1) in the form of the random arrangement of internal partial optical elements. ) Can be generated. That is, the basic optical element P (j) is subjected to the random arrangement according to the j-th form different from the first form of the random arrangement used when the divided basic optical element P (1) is generated.
Can be generated. In the present embodiment, the generation operation of the divided basic optical element described above is performed 30 times while changing the form of the random array, and 30 basic optical elements P (1) to P (P) to P (1) to P
(30) are formed, and these 30 basic optical elements P are formed.
A pattern of the diffractive optical element is formed by integrating (1) to P (30) so that the same number is included in the effective diameter of the diffractive optical element for each type j.

FIG. 78 is a diagram showing a simulation result of the light intensity distribution when the diffractive optical element of the eleventh embodiment is applied to the exposure apparatus of FIG. That is, the simulation result shown in FIG. 78 is obtained by arranging the diffractive optical element of this embodiment as the diffractive optical element 21 in the exposure apparatus of FIG.
The illumination distribution (light intensity distribution) formed by the diffractive optical element on the incident surface of the fly-eye integrator 6 via the relay lens system 4 when the diffractive optical element is incident on the diffractive optical element after being shaped into a rectangular section of 0 mm. ). In addition, on the illumination pupil plane near the exit surface of the fly-eye integrator 6, the illumination distribution (light intensity distribution) formed on the entrance surface
An illumination distribution (light intensity distribution) that substantially corresponds to is formed.

Referring to FIG. 78, in the eleventh embodiment, it is possible to form a uniform light intensity distribution on the illumination pupil plane, which is similar to the outer shape of the divided projection basic region of FIG. 77 (a). Recognize. In particular, in the present embodiment, most of the partial optical elements are set to have a symmetrical shape in the x direction and the y direction, that is, a square shape, so that the sharpness of the edge of the illuminance distribution obtained on the wafer conjugate plane and the illumination pupil plane is small. Almost uniform in the x direction and the y direction, and good imaging performance and optical performance can be secured. Further, as described above, since the phase type binary Fresnel zone plate is used as the predetermined optical element (virtual lens),
It is actually possible to form the entire ring-shaped basic illumination area by using only the divided basic optical elements P (1) to P (30) corresponding to the upper half divided projection basic areas.

In the eleventh embodiment, as in the modification of the seventh embodiment described above, in addition to the divided basic optical element P (j), the divided basic optics obtained by rotating 90 degrees in the xy plane. It is preferable to introduce the element P (j) '_ 90. In this case, the basic optical element P (j) designed by the quantization along the y direction and the basic optical element P (j) ′ _ 90 designed by the quantization along the x direction are defined as the effective diameter of the diffractive optical element. By mixing substantially the same number in the inside, it is possible to form illumination without quantization anisotropy, that is, a ring-shaped basic illumination area having a good circular approximation in the illumination pupil plane.

Further, in the eleventh embodiment, the diffractive optical element of the present embodiment is applied to the diffractive optical element 21 in the exposure apparatus of FIG. 7 to perform annular illumination having an annular ratio close to 1. However, without being limited to this, for example, the diffractive optical element of the present embodiment is applied to the diffractive optical element 6 in the exposure apparatus (see FIG. 1) disclosed in JP 2001-85293 A. Band illumination can also be performed. Here, the diffractive optical element 6 in the publication has a ring-shaped (a narrow annular zone having an annular zone ratio close to 1) light intensity distribution in the far field when a parallel light flux having a rectangular cross section is incident. Has the function of forming. As described above, the diffractive optical element of the present embodiment can be applied to the diffractive optical element having the same function as the diffractive optical element 6 in the publication, regardless of the purpose.

By the way, in the exposure apparatus of FIG. 7, the relay lens system 4 connects the diffractive optical element 21 and the incident surface of the fly-eye integrator 6 in a substantially Fourier transform relationship. However, in each of the above-described embodiments and modifications, in order to perform illumination with higher uniformity on the wafer conjugate plane and the illumination pupil plane, the diffractive optical element 21 via the relay lens system 4 side to the fly-eye integrator 6 side. The incident surface of the fly-eye integrator 6 may be displaced from the Fourier transform surface so that the light flux that converges on the input surface of the fly-eye integrator 6 is slightly blurred due to aberration or the like.
Specifically, the blur width can be set to a desired value by appropriately selecting the focal length of the lens on the diffractive optical element 21 side of the relay lens system 4. For example 0.1 mm
By setting the blur width value to about 5 mm, it is possible to perform illumination with higher uniformity on the wafer conjugate plane and the illumination pupil plane.

Further, in each of the above-described embodiments and modified examples, when the size of the basic optical element is fixed, if the division into the partial optical elements is made too coarse (that is, the size of the partial optical element is set large). If too much), the dividing line of the partial optical element may be slightly visible, so the number of divisions into the partial optical element should not be made too rough. Conversely, if the division into the partial optical elements is made too fine (that is, if the size of the partial optical elements is set too small), the sharpness of the edges may be impaired in the light intensity distribution of the illumination area, so the partial optical elements may be damaged. Do not make the number of divisions into elements too small.

The optimum value of the size of the partial optical element is
It also depends on variables such as the wavelength used, the focal length of the relay lens, and the focal length of the virtual lens. Further, the influence of the division line generated when the division of the partial optical element is slightly rough can be averaged by the aberration of the relay lens system 4. The optimum value of the size of the partial optical element and the value of the aberration of the relay lens are finally computer-simulated with respect to each embodiment including the diffractive optical element (basic optical element), and the illumination unevenness on the reticle conjugate plane and the pupil plane is calculated. It is desirable to set so that the optical performance such as the above satisfies the specifications. For example, in the above-described modified example of the sixth embodiment, the size of the equally divided partial optical element is set to 40 μm × 40 μm, the blur width due to aberration is set to about 2 mm, and it is confirmed that good illumination performance can be achieved. There is. A computer simulation is also performed for each of the modified example of the seventh embodiment and the ninth to eleventh embodiments, and the size of the partial optical element and the blur width of the aberration are appropriately set.

In each of the above-described embodiments and each modification, the phase type binary Fresnel zone plate, that is, the binary phase type Fresnel zone plate is used as the predetermined optical element (virtual lens), but it is not limited to this. Without mentioning, for example, in JP 2001-174615 A, the blazed type phase zone plate described in reference to FIG. 14 (blaze type phase Fresnel zone plate), a multi-stage binary phase zone plate with three or more stages (multivalued A phase type Fresnel zone plate), a refraction type microlens, or the like can be used. In this case, since the diffracted light (or refracted light) is not generated symmetrically, in the modified example of the sixth embodiment, the modified example of the seventh embodiment, and the ninth to eleventh embodiments, the divided basic optical element P is used.
It is necessary to configure a basic optical element corresponding to each j by arranging (j) and a divided basic optical element P (j) d obtained by vertically inverting it with respect to the x-axis in one set. Needless to say.

As described above, the binary phase type Fresnel zone plate is the same as what is called the phase type binary Fresnel zone plate. However, more specifically, the phase-type Fresnel zone plate including the binary type is generally applicable to L-value quantized L-value (L ≧ 2) phase-type Fresnel zone plate. The name of binary phase type Fresnel zone plate is used by displaying up to the number of steps. Hereinafter, in connection with the above, when the virtual lens used in the present invention is a blazed phase Fresnel zone plate, an L-value phase Fresnel zone plate, and a binary phase-type Fresnel zone plate limited to L = 2. The case will be further described with reference to FIG. 79.

FIG. 79 (a) is a sectional view of a blazed phase Fresnel zone plate taken along a section including the center of rotation of the zone plate pattern, and FIG. 79 (b) is an L-value phase Fresnel zone plate. FIG. 8 is a cross-sectional view taken along a cross section including the center of rotation of the pattern.
9C is a cross-sectional view of the binary phase type Fresnel zone plate taken along a section including the center of rotation of the zone plate pattern. Referring to FIG. 79 (a), the cutting surface of the blazed phase Fresnel zone plate has a sawtooth shape, and the radius r _m of the ring that defines the position on the concentric circle where the step occurs in a sawtooth shape is given by the following equation. It is given in (14). r _m = {(m · λ) ² + 2 · m · f · λ} ^1/2 (14)

Here, λ is the wavelength of light, f is the focal length of the virtual diffraction lens, and m is an arbitrary integer. A cross section of a concentric zone plate pattern defined by the radius r _m is shown in FIG. 79 (a). The depth d ₀ in the cross section is given by the following equation (15), where n is the refractive index of the substrate and 1 is the refractive index of the gas in which the substrate is arranged. d ₀ = λ / (n-1) (15)

By using such a blazed type phase Fresnel zone plate as a virtual lens, it is possible to determine the partial optical element and the basic optical element by the method of the present invention. In this case, the patterns of the rearranged partial optical elements and basic optical elements are not the binary type as shown in FIGS. 68 and 69, but the cross-sectional shape changes in the height direction. When generating such a pattern having different height directions, it is possible to use a gray scale mask whose transmittance gradually changes. That is, a grayscale mask pattern for manufacturing a rotationally symmetric blazed phase zone plate by exposure and etching is prepared in advance as a virtual lens pattern, and the present invention is applied to the grayscale mask pattern of the virtual lens. The setting method of the partial optical element and the basic optical element can be applied.

Using the gray scale mask pattern for the basic optical element prepared as described above, pattern transfer and etching are performed on a transparent substrate such as synthetic quartz glass coated with a resist to obtain a rotationally symmetric blazed type. It is possible to manufacture a partial optical element having a pattern obtained by cutting out a predetermined portion of the phase Fresnel zone plate of, and a basic optical element in which the partial optical element is densely rearranged.

Referring to FIG. 79 (b), the cut surface of the L-value phase type Fresnel zone plate has a shape similar to the saw-tooth shape of FIG. 79 (a) like an L-step staircase shape. Then, the ring radius r _k indicating the boundary of each stair can be easily defined by dividing the range of r _m and r _{m + 1} shown in Expression (14) according to the step area of each stair. FIG. 79B shows the case of an 8-value phase type Fresnel zone plate with L = 8. The depth d _L in the cross section is given by the following equation (16), where n is the refractive index of the substrate and 1 is the refractive index of the gas in which the substrate is arranged. d _L = λ · (L−1) / {L · (n−1)} (16)

By using such an L-value phase type Fresnel zone plate as a virtual lens, the partial optical element and the basic optical element can be determined by the method of the present invention.
In this case, the patterns of the rearranged partial optical elements and basic optical elements are 2 as shown in FIGS. 68 and 69.
Instead of the value type, the cross-sectional shape becomes a staircase shape, and the pattern changes in the height direction. When generating such a pattern having different height directions, it is possible to use a gray scale mask whose transmittance changes stepwise and in multiple stages. That is, a grayscale mask pattern for manufacturing a rotationally symmetric L-value phase type Fresnel zone plate by exposure and etching is prepared in advance as a virtual lens pattern, and the grayscale mask pattern of the virtual lens is prepared as a main pattern. The partial optical element and basic optical element setting methods of the invention can be applied.

By using the gray scale mask pattern for the basic optical element prepared as described above, pattern transfer and etching are performed on a transparent substrate such as synthetic quartz glass coated with a resist to obtain a rotationally symmetric L value. It is possible to manufacture a partial optical element having a pattern obtained by cutting out a predetermined portion of the phase type Fresnel zone plate, and a basic optical element in which the partial optical elements are densely rearranged.

As another method not using the gray scale mask, a black-and-white type mask (only the transmitting portion and the light shielding portion) that selectively patterns only the range of each staircase is used a plurality of times, and each step of each staircase is used. A method of patterning for each pattern is also possible. In this case, a plurality of black-and-white masks for manufacturing a rotationally symmetric L-value phase Fresnel zone plate by exposure and etching are prepared in advance as patterns of virtual lenses, and each black-and-white mask for the virtual lenses is prepared. The method of setting the partial optical element and the basic optical element of the present invention can be applied to the mask pattern.

Using each black-and-white type mask pattern for the basic optical element thus created, a transparent substrate such as synthetic quartz glass coated with a resist is subjected to a plurality of exposure operations and etchings to obtain rotational symmetry. It is possible to manufacture a partial optical element having a pattern obtained by cutting out a predetermined portion of the L-value phase type Fresnel zone plate, and a basic optical element in which the partial optical elements are densely rearranged.

Referring to FIG. 79 (c), the cross section of the binary phase type Fresnel zone plate (same as the phase type binary Fresnel zone plate) is L = 2 in the step type approximation of FIG. 79 (b). It has a shape of. That is, as shown in FIG. 79 (c), the cut surface of the binary phase type Fresnel zone plate has a rectangular shape shown by two kinds of steps of a convex region portion and a concave region portion in the direction of the thickness d ₂ . It becomes a pattern.

Ring radius r indicating the boundary of each rectangular area
_m can be given by the aforementioned equation (1) already described. The depth d ₂ in the cross section is given by the following equation (17), where n is the refractive index of the substrate and 1 is the refractive index of the gas in which the substrate is arranged. d ₂ = λ / {2 · (n −1)} (17)

By using such a binary phase type Fresnel zone plate as a virtual lens, it is possible to determine the partial optical element and the basic optical element by the method of the present invention.
That is, a pattern of a black-and-white type mask (only a transmission part and a light shielding part) for manufacturing a rotationally symmetric binary phase type Fresnel zone plate by exposure and etching is prepared in advance as a virtual lens pattern, and the virtual lens is prepared. The setting method of the partial optical element and the basic optical element of the present invention can be applied to the pattern.

Using the binary phase type Fresnel zone plate pattern for the basic optical element thus prepared,
A partial optical element having a pattern obtained by cutting a predetermined portion of a rotationally symmetric binary phase type Fresnel zone plate by performing pattern transfer and etching on a transparent substrate such as synthetic quartz glass coated with a resist, and the partial optical element. As described in the above embodiments, it is possible to manufacture a basic optical element in which elements are densely rearranged.

68 and 69 described above, a binary phase type Fresnel zone plate is applied just as a virtual lens, and the patterning rule of the partial optical element and the basic optical element of the present invention is applied to the pattern of the virtual lens. Is applied, and a mask pattern created by the patterning rule is transferred to form a basic optical element on synthetic quartz glass. That is, in the figure, the white portion of the pattern is a binary phase type Fresnel zone plate in which the white area indicates the convex area portion and the black portion indicates the concave area portion, and the black portion represents the light shielding. I'm not. In other words, in the figure, the white part and the black part merely represent the step difference in the d ₂ direction of the pattern patterned on the glass substrate.

Further, in the modification of the sixth embodiment, the modification of the seventh embodiment, and the ninth to eleventh embodiments, the equally-divided partial area has a size qx in the x direction and a size in the y direction. The length qy is set to be equal. However, the present invention is not limited to this, and the size qx in the x direction and the size qy in the y direction of the equally-divided partial regions are finally adjusted so that the illuminance distributions on the wafer conjugate plane and the illumination pupil plane become optimal.
Is preferably set to an optimum value.

Further, a seventh modified example of the above-described sixth embodiment.
In the modified examples of the embodiment, the ninth embodiment to the eleventh embodiment, the partial optical elements are 5 μm × 5 μm to 300 μm × 30.
It is preferable to set a substantially square shape of 0 μm. In addition, it is preferable to set the number of pitches of the patterns included in the partial optical elements (partial optical elements near the minimum pitch) farthest from the optical axis before rearrangement to 3 pitches or more. Furthermore, the basic optical element is 50 μm × 50 μm
It is preferable to set so that the area is about 2000 μm × 2000 μm. Further, it is desirable to set the focal length of the relay lens system 4 to 100 mm or more.
Furthermore, the number of partial optical elements included in the basic optical element is 5
It is preferable to set it to about 0 to 5000 pieces. Also,
It is preferable to set the number of independent basic optical elements integrated within the effective diameter of the diffractive optical element to 4 or more (more preferably 20 or more).

By the way, in each of the above-described embodiments and modifications, a phase type binary Fresnel zone plate defined by the following formula (18) may be used instead of the above formula (1) or (14). it can. Further, instead of the above-mentioned formula (1), a formula expressing a phase distribution corresponding to an aspherical surface such that the illuminance distribution on the wafer conjugate plane and the illumination pupil plane finally becomes optimal can be used. r _m = (m · f · λ) ^1/2 (18)

Furthermore, in each of the above-described embodiments and modifications, the illuminance distributions on the wafer conjugate plane and the illumination pupil plane are finally optimal, as described in paragraph 0085 and subsequent paragraphs of Japanese Patent Laid-Open No. 2001-174615. It is also possible to correct the phase distribution of the partial optical element so that

The present invention can be applied to create a far field distribution having a predetermined shape without using the relay lens system 4 or the like. In this case,
In designing the basic optical element, a relay lens having a very long focal length (for example, a focal length of 1000 mm or more) can be virtually applied and designed. Then, in actual use, the relay lens is removed and used. Further, it is also possible to form the predetermined basic optical element for each of the three colors of RBG and apply the present invention as a color display application having a predetermined shape. In this case, it is preferable to use a multi-valued phase type Fresnel zone plate having four or more stages or a blazed type phase zone plate as the predetermined optical element (virtual lens) for generating the diffractive optical element.

FIG. 80 is a diagram schematically showing another configuration of an exposure apparatus to which the diffractive optical element according to each embodiment and each modification of the present invention can be applied. The exposure apparatus shown in FIG. 80 has a configuration similar to that of the exposure apparatus shown in FIG. 7. However, in the exposure apparatus of FIG. 80, an optical delay system MB and a reduction are provided in the optical path between the beam shaping optical system 2 and the diffractive optical element 21. 7 in that a beam shaping optical system 113 of a mold and a depolarizing prism system 114 are additionally provided. The exposure apparatus of FIG. 80 will be described below, focusing on the differences from the exposure apparatus of FIG.

In the exposure apparatus of FIG. 80, the light beam converted into the desired cross-sectional shape through the beam shaping optical system 2 enters the optical delay system MB composed of the total reflection mirror 111 and the partial reflection mirror 112. . In the optical delay system MB, a part of the light flux transmitted through the partial reflection mirror 112 enters the beam shaping optical system 113, and the remaining light flux reflected by the partial reflection mirror 112 enters the total reflection mirror 111. The light flux reflected by the total reflection mirror 111 is incident on the partial reflection mirror 112, and a part of the light flux transmitted through the partial reflection mirror 112 is incident on the beam shaping optical system 113, so that the partial reflection mirror 11
The remaining light flux reflected by 2 is incident on the total reflection mirror 111.

In this way, in the optical delay system MB, a large number of beam groups in which the incident beams are sequentially delayed by the multiple reflections repeated between the total reflection mirror 111 and the partial reflection mirror 112 (for simplification in FIG. 80). (Only the five beams are schematically shown). As a result, the optical delay system MB
By the action of, the interference noise on the wafer conjugate plane can be reduced. For more detailed structure and operation of the optical delay system MB, see, for example, Japanese Patent Laid-Open No. 9-205.
Reference can be made to JP-A No. 060, JP-A-10-125585, and JP-A-2000-277421.

Beam shaping optical system 1 via optical delay system MB
The light flux entering 13 is reduced and converted into a desired cross-sectional shape, and then enters the depolarizing prism system 114. The depolarization prism system 114 is composed of, for example, a wedge prism formed of quartz and a wedge prism formed of synthetic quartz. It has a function of converting into a light flux of. As a result, the depolarizing prism system 114
By the action of, the isotropy of the pattern image formed on the wafer surface is improved. It should be noted that a more detailed configuration and action of the depolarizing prism system 114 will be described in, for example, Japanese Patent No.
No. 679319 and No. 2679337 (and corresponding US Pat. No. 5,253,11).
No. 0) can be referred to. When only the depolarization prism system is applied to the exposure apparatus shown in FIG. 7, for example, the depolarization prism system may be provided in the optical path between the beam shaping optical system 2 and the diffractive optical element 21.

FIG. 81 is a diagram schematically showing still another configuration of the exposure apparatus to which the diffractive optical element according to each embodiment and each modification of the present invention can be applied. Also, FIG.
FIG. 82 is a diagram schematically showing a configuration of a main part of the exposure apparatus of FIG. 81.
The exposure apparatus shown in FIG. 81 has a configuration similar to that of the exposure apparatus shown in FIG. 80. However, in the exposure apparatus of FIG. 81, an internal reflection type optical integrator is used in place of the fly-eye integrator 6 which is a wavefront division type optical integrator. The difference from the exposure apparatus of FIG. 80 is that a certain rod integrator (rod type integrator) 121 is used. The exposure apparatus of FIG. 81 will be described below, focusing on the differences from the exposure apparatus of FIG.

In the exposure apparatus of FIG. 81, the rod integrator 121 is arranged in place of the fly-eye integrator 6, so that an input is made in the optical path between the relay lens system 4 and the rod integrator 121. The lens 122 is arranged and the rod integrator 121 is arranged.
A relay lens 123 is arranged in the optical path between the condenser lens system 8 and the condenser lens system 8. Here, FIG. 81 and FIG.
The surface A shown in FIG. 8 corresponds to the incident surface of the fly-eye integrator 6 in FIG. 80, and the surface B shown in FIGS. 81 and 82 is shown in FIG.
It corresponds to the exit surface of the fly-eye integrator 6 of 0.

The rod integrator 121 is an internal reflection type glass rod made of a glass material such as quartz glass or fluorite, and utilizes total internal reflection on the boundary surface between the inside and the outside, that is, the inside surface to collect the light. A number of light sources corresponding to the number of internal reflections are formed along a plane parallel to the rod incidence plane. Here, most of the light sources formed are virtual images, but only the light source at the center (condensing point) is a real image. That is, the light beam incident on the rod integrator 121 is angularly divided by internal reflection, and a secondary light source composed of a large number of light sources is formed along a plane that passes through the converging point and is parallel to the incident plane.

Therefore, the light flux passing through the diffractive optical element 21 has a predetermined shape on the surface A (a ring shape, a quadrupole shape, a circular shape, etc.).
After the illumination field is formed, the light is condensed in the vicinity of the incident surface 121a of the rod integrator 121 via the input lens 122. Thus, rod integrator 1
21 has a predetermined shape (annular shape,
The light flux from the secondary light source having a quadrupole shape, a circular shape, or the like is superposed on the exit surface 121b, and then the relay lens 12
A reticle (mask) 9 on which a predetermined pattern is formed is superposedly illuminated via 3 and a condenser lens system 8. It is also possible to remove the relay lens 123 and the condenser lens system 8 and install the exit surface 121b of the rod integrator 121 near the reticle 9.

FIG. 83 shows the relationship between the fly-eye integrator having a configuration different from that of the fly-eye integrator of FIG. 7 and the annular opening of the annular aperture stop, and the annular shape to be formed on the incident surface. It is a figure which shows the relationship between the illumination area of FIG. In the above description, as shown in FIG. 9, in two rows adjacent to each other in the y-direction (scanning direction), the fly-eyes in which the pitches of the lens elements EL along the x-direction are displaced by 1/2 pitch are arranged. Although the present invention has been described based on the integrator 6, instead of the fly-eye integrator 6, FIG.
It is also possible to use a fly-eye integrator 6d in which the lens elements EL are aligned vertically and horizontally as shown in FIG.

In FIG. 83, the fly-eye integrator 6d and the ring-shaped transmission area AP500 of the ring-shaped aperture stop arranged on its exit surface are shown in an overlapping manner. The rectangular area SP displayed inside each rectangular lens element EL, which is close to a square to some extent, indicates a light spot area. In the figure, it is assumed that the filling rate of the light spot region SP is 50%. The lens element EL including the light spot area SP in the ring-shaped transmission area AP500 is a necessary lens element.

FIG. 84 is a diagram showing how a semi-zonal split basic illumination area is set according to the necessary lens elements defined in FIG. Referring to FIG. 84, a divided basic illumination area IA500 (indicated by a thick line in the drawing) for the area of y ≧ 0 according to the necessary lens elements defined in FIG. 83.
However, in practice, a region in which a slight margin portion is added to the outside of the region defined by the necessary lens elements is set as the divided basic illumination region IA500.

Flyeye integrator 6d
Also in the case of using, the divided basic illumination area IA500 is projected on the virtual lens and the divided projection basic area PIA500 (not shown) is set, as in the above-described respective embodiments and modified examples. Further, the divided projection basic area PIA500 is divided along the y direction at a pitch of qy, and divided into a large number of partial areas extending along the x direction. Further, dividing lines are set in a large number of partial areas to define a large number of partial areas, and thus a large number of partial optical elements, and these basic optical elements K50 are densely rearranged in a rectangular shape.
A pattern of 0 can be determined.

For example, similar to the modification of the sixth embodiment, it is possible to further subdivide the partial optical elements and arrange them randomly, and in subdividing, the size qx in the x direction is the size in the y direction. It can be set to be almost the same as qy. When a phase type binary Fresnel zone plate is assumed as the virtual lens, the divided basic illumination area IA50 for the half area (y ≧ 0).
The basic optical element K500 defined from 0 also illuminates an area in which the divided basic illumination area IA500 is folded back symmetrically with respect to the x-axis. That is, the basic optical element K500 can secure an illumination area required for annular illumination.

In FIG. 85, basic optical elements of different types having different internal arrangements generated based on the divided basic illumination area defined in FIG. 84 are patterned so that the same number is included in the effective diameter of the diffractive optical element. FIG. Referring to FIG. 85, similar to the modification of the sixth embodiment, for example, a plurality of basic optical elements K500 having different arrangement forms of subdivided partial optical elements are prepared, and each type of basic optical element K500 (j) is provided. The diffractive optical elements D500 are arranged so that the same number is included in the effective diameter FD500.

In the above description, the present invention has been described with reference to the scan exposure type exposure apparatus.
The diffractive optical element of the present invention can also be applied to a collective exposure type exposure apparatus. The batch exposure type exposure apparatus is an exposure apparatus that collectively exposes a reticle pattern to each exposure area of a wafer without performing scan exposure, that is, performs a batch exposure in a certain exposure area, and then performs the next exposure after the exposure to that exposure area. The exposure apparatus repeats the exposure operation of stepwise moving to the exposure area and performing different batch exposure. In such a batch exposure type exposure apparatus, each exposure region exposed by one batch exposure is often set in a substantially square shape.

The lens element EL of the fly-eye integrator (or microlens array) is almost similar to the illumination area on the reticle, and thus to each exposure area on the wafer, so that the fly-type exposure apparatus is used. The lens element EL of the eye integrator needs to have a substantially square shape. That is, the fly-eye integrator 6d in FIG. 83 can be applied to a collective exposure type exposure apparatus. Further, the effective diameter FD500 of the diffractive optical element D500 in the preceding stage is generally imaged in the light spot region SP. Therefore, in order to fill the light spot region SP in each lens element EL as closely as possible, the effective diameter FD5 of the diffractive optical element D500 is set.
It is preferable to set the shape of 00 to a substantially square shape like the lens element EL.

It should be noted that FIG. 8 shows a single exposure type exposure apparatus.
Vertically and horizontally aligned fly-eye integrator 6d shown in 3
The example of applying the above is described. However, a collective exposure type exposure apparatus using a fly-eye integrator obtained by alternately arranging lens elements EL having a substantially square shape as shown in FIG. 83 is implemented. Of course, it is also possible. Further, as described above, instead of the fly-eye integrator, a fine lens array group is formed on the glass substrate, and a batch exposure using a micro-lens array that has the same effect as the fly-eye integrator is miniaturized. Embodiments of mold type exposure apparatus are also possible.

In this case, the basic illumination area is not set based on the light spot area SP of the lens element EL as shown in FIG. 84, but a ring-shaped pupil distribution as in the seventh embodiment and its modification. It is possible to set a basic illumination area that directly forms the light source and design the diffractive optical element of the present invention based on this basic illumination area. Further, it is of course possible to apply the fly-eye integrator obtained by aligning the rectangular lens elements EL as shown in FIG. 9 to the scan exposure type exposure apparatus as shown in FIG.

Since the pupil distribution of illumination is directly formed by the diffractive optical element of the present invention in the seventh and eighth embodiments and its modified examples, the basic illumination area of the diffractive optical element is geometric. It is preferable to set the shape so as to be as close as possible to the predetermined pupil shape at the stage. This can be performed by adjusting the division pitch of the partial optical elements, that is, the division pitch qy when the basic illumination area is quantized at equal intervals in the y direction by a straight line parallel to the x axis. That is, when the value of the division pitch qy is changed, the quantization state (that is, the outer shape defined by the quantization) changes at the upper and lower boundaries in the y direction in the basic illumination area or the projection basic area.

For example, in the projection basic area PIA301 shown in FIG. 63 in relation to the modification of the seventh embodiment, qy (= qx) = on the incident surface of the fly-eye integrator 6.
This is the result when the distance is 1.00 mm. On the other hand, the projection basic area PIA700 shown in FIG. 86 has qy (= qx) = 0.97 on the incident surface of the fly-eye integrator 6.
9 mm, and other conditions for PIA301 in FIG.
The result is the same as. Referring to FIG. 86,
In the projection basic area PIA700, the smoothness is slightly worse in the upper part in the y direction.

As described above, only by changing qy (= qx) by about several percent, a large difference occurs in the outer shape between the projection basic areas PIA301 and PIA700. Therefore, the results of the two-dimensional diagram of the basic illumination area or the projection basic area in which the value of the division pitch qy is changed within an appropriate range are listed in advance, and the results are smoothed like the projection basic area PIA301 shown in FIG. 63. From the viewpoint of size, it is desirable to select the basic illumination area having a preferable outer shape, and thus the projection basic area.

Here, note the qy and qx used in the description of the present invention with reference to FIG. The principle of patterning is the same in the case of considering the incident surface of the fly-eye integrator 6 or a corresponding surface thereof and the case of considering the surface corresponding to the virtual lens (or the predetermined optical element or the diffractive optical element). It is possible to define quantities corresponding to qy and qx also in the plane of, but as the actual values of qy and qx, in the sense of FIG. ) Need to think. On the contrary, except for the magnification conversion amount, it is conceptually the same in any aspect, and therefore the same symbols may be used in both aspects as symbols for qy and qx.

As described above, instead of the fly-eye integrator, a fine lens array group is formed on the glass substrate, and a micro-lens array having the same effect as that of the fly-eye integrator is miniaturized. Also, an embodiment of a single exposure type exposure apparatus is possible. Also,
In both the scan exposure type and the batch exposure type exposure apparatuses, a type in which a pair of one-dimensional cylindrical lens arrays are arranged so as to be orthogonal to each other is not a simple microlens array formed of a normal microlens group. A microlens array can also be used.

Further, as a substrate for patterning the diffractive optical element according to the present invention, a transparent material such as ordinary glass material, plastic, synthetic quartz, crystal, fluorite can be used without particular limitation.

Further, for the purpose of obtaining a better pupil plane distribution and wafer surface illumination distribution, the pattern of the basic optical element of the present invention described above is used as an initial pattern, and an optimization method is applied to the initial pattern. By applying it, it is possible to generate an optimized basic optical element.
For example, the basic optical element in the modification of the seventh embodiment has a random arrangement as shown in FIG. 63 (b).
Further, this basic optical element has an internal pattern as shown in FIG. With this basic optical element pattern as an initial pattern, various optimization algorithms can be applied.

The optimization algorithm is an algorithm for repeating the calculation based on a certain initial value and sequentially converging the initial value to the ideal value. As an optimization algorithm, an iterative method (author: Michael R. Feldman et al.,
Title, Source: Iterative Encoding of High Efficiency
Holograms for Generation of Spot Arrays, OpticsLe
tters, Vol. 14, pp. 479-81, 1989), simulated annealing, GA (genetic algorithm: genetic)
algorithm) is generally known. However, the method is not limited to these methods, and any algorithm that makes successive improvements using the initial distribution can be used.

As a method of applying the present invention, for example, regarding an arrangement method of basic optical elements, an optimization algorithm for optimizing the arrangement for a random arrangement pattern shown in FIG. It may be applied to obtain the optimum sequence. Here, the optimum array means a basic optical element array in which the pupil plane distribution and the wafer plane illumination distribution are improved as compared with the initial pattern. The procedure for manufacturing a basic optical element using this optimized pattern is the same as described above.

Further, the partial optical element shown in FIG. 63 (b), FIG. 68, etc. is not used as it is, but the phase distribution of a part or the whole of the basic optical element is placed on the mesh set on the xy plane. The expressed initial phase field φ0 (x, y) is set, and this phase field φ0 (x, y) is used as an initial pattern.
There is also an application method in which an optimization algorithm for optimizing the phase distribution is applied to obtain the optimum phase field φ (x, y). Here, the optimum phase field means a phase field in which the pupil plane distribution and the wafer plane illumination distribution are improved as compared with the initial phase field, and the optimum phase field or a plurality of optimized phase fields are It plays a role similar to the basic optical element of the invention.

Such an optimal phase distribution of the phase field φ (x, y) is represented by a blazed type or an L value phase distribution that approximates it, and as described above, a gray scale mask or one or more An optimized basic optical element is manufactured by forming a pattern based on the phase field on a black and white mask, and transferring and etching the pattern on a transparent substrate such as synthetic quartz glass coated with resist using the mask. It is possible to Further, the optimization algorithm is applied only to the arrangement method of the partial optical elements in advance, and the initial phase field φ0 (x, y) similar to the above is set based on the basic optical element whose arrangement is optimized, and the optimization is performed for this. Applying the algorithm, the optimum phase field φ
It is also possible to obtain (x, y).

As described above, the illuminance on the wafer conjugate plane or the illumination pupil plane is applied by applying, for example, an optimization algorithm with the pattern of the basic optical element obtained in each of the above-described embodiments and modifications as the initial distribution. The pattern of the basic optical element can be optimized so that the distribution or shape has a desired illuminance distribution or a desired shape. That is, by using the array of partial optical elements in the basic optical element obtained in each of the above-described embodiments and modified examples as an initial value and optimizing the array of partial optical elements using an optimization algorithm, the wafer conjugate plane At least one of the illuminance distribution on the illumination pupil plane and the illuminance distribution on the illumination pupil plane can be brought into a desired state.

Also, the array of partial optical elements and the internal pattern in the basic optical element obtained in each of the above-described embodiments and modifications are used as initial values to further subdivide the internal pattern of each partial optical element, and the optimization algorithm To optimize the arrangement of the partial optical elements and the arrangement of the subdivided internal patterns, it is possible to bring at least one of the illuminance distribution on the wafer conjugate plane and the illuminance distribution on the illumination pupil plane to a desired state. it can.

As the light source 1 in each of the above embodiments, a KrF excimer laser for supplying a laser beam having a wavelength of 248 nm and an A for supplying a laser beam having a wavelength of 193 nm are used.
An rF excimer laser, an F ₂ excimer laser that supplies laser light having a wavelength of 157 nm, a harmonic of a YAG laser, or an i-line (wavelength 365 nm) of a mercury lamp can be used.

Further, the application of the projection exposure apparatus according to each of the above-mentioned embodiments is not limited to the exposure apparatus for semiconductor manufacturing, and for example, for exposing a liquid crystal display element pattern on a rectangular glass plate. Exposure equipment for liquid crystal,
It can be widely applied to an exposure apparatus for manufacturing a thin film magnetic head.

Further, as the projection optical system of each of the above-mentioned embodiments, a refraction type projection optical system, a reflection type projection optical system, a catadioptric type projection optical system, etc. can be applied. The projection magnification of the projection optical system is not limited to the reduction magnification, and may be the same magnification or the enlargement magnification. As described above, various modifications of the present invention are possible.

[0364]

As described above, in the optical element according to the present invention, a plurality of densely arranged partial optical elements convert an incident light beam into a plurality of partial light beams, and a plurality of densely arranged partial optical elements. The basic optical element, which is the sum of the elements, converts the incident light beam into a light beam having a predetermined light beam cross section as the total of a plurality of partial light beams. As a result, with the optical element of the present invention, it is possible to efficiently convert the incident light beam into a light beam having a predetermined cross-sectional shape (while suppressing the light amount loss satisfactorily).

In particular, in the present invention, the partial optical elements are subdivided and randomized, and a large number of random combinations are performed to average out illumination unevenness due to diffraction unevenness and manufacturing errors, thereby achieving uniform illumination on the wafer conjugate plane. Also, the illumination uniformity (σ uniformity or pupil distribution uniformity) in the illumination pupil can be significantly improved.

Further, in the illumination optical device provided with the optical element of the present invention, by switching a plurality of optical elements, it is possible to form various light intensity distributions while satisfactorily suppressing the light amount loss in the illumination pupil. Similarly, in an exposure apparatus equipped with an illumination optical apparatus including the optical element of the present invention, good projection exposure is performed and good devices are manufactured based on various modified illuminations in which light quantity loss is favorably suppressed. can do.

[Brief description of drawings]

FIG. 1 is a flowchart showing a manufacturing flow of a mask used for manufacturing an optical element according to the present invention by lithography.

FIG. 2 is a diagram illustrating a first dividing step in the present invention.

FIG. 3 is a diagram illustrating a total length calculation step in the present invention.

FIG. 4 is a diagram illustrating a second dividing step in the present invention.

FIG. 5 is a diagram illustrating a third dividing step in the present invention.

FIG. 6 is a diagram illustrating a dense arrangement step in the present invention.

FIG. 7 is a diagram schematically showing a configuration of an exposure apparatus equipped with an illumination optical device including a diffractive optical element according to each embodiment of the present invention.

8 is a diagram showing a plurality of diffractive optical elements provided in a first revolver and a plurality of aperture stops provided in a second revolver in FIG. 7. FIG.

FIG. 9 is a diagram showing the relationship between the fly-eye integrator and the ring-shaped aperture of the ring-shaped aperture stop in the first embodiment.

FIG. 10 is a diagram showing a relationship between an annular illumination region to be formed on the incident surface of the fly-eye integrator and a large number of lens elements in the first embodiment.

FIG. 11 is a diagram showing a projected basic area obtained by projecting a basic illumination area onto a predetermined virtual diffraction lens.

FIG. 12 is a diagram illustrating a relationship between a virtual optical lens virtually arranged at the position of a diffractive optical element and a basic optical element.

13A is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions, and FIG. 13B is defined corresponding to the large number of partial regions divided in FIG. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging a large number of partial optical elements densely.

FIG. 14 is a diagram illustrating a division rule in FIG.

15A is a diagram showing a state in which the lower half divided projection basic region is divided into a large number of partial regions, and FIG. 15B is defined corresponding to the large number of partial regions divided in FIG. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging a large number of partial optical elements densely.

16 is a split basic optical element of FIG. 13 (b) and FIG.
It is a figure which shows a mode that a basic optical element is formed by composition with the division | segmentation basic optical element of (b).

FIG. 17 is a diagram showing a state in which the center of the projection basic area and the center of the FZP pattern are eccentric.

FIG. 18 is a diagram showing an arrangement of four types of basic optical element patterns formed on a mask used for manufacturing a diffractive optical element including four types of basic optical elements in which the positions of basic illumination regions are slightly different from each other. .

FIG. 19 is a diagram illustrating application of a vertical inversion rule, a horizontal inversion rule, and an up / down / left / right inversion rule to a basic optical element pattern.

20A is a diagram showing an arrangement of 9 types of basic optical elements when the number of overlaps is 9, and FIG. 20B is a diagram showing an arrangement of 16 types of basic optical elements when the number of overlaps is 16. FIG. Is.

FIG. 21 is a diagram schematically showing a configuration of a mask used for manufacturing the diffractive optical element according to the first embodiment by lithography.

22 is a diagram showing one diffractive optical element of a plurality of diffractive optical elements formed on a glass substrate using the mask of FIG. 21.

FIG. 23 is a diagram showing the relationship between a circular illumination region to be formed on the incident surface of the fly-eye integrator and a large number of lens elements in the second embodiment.

FIG. 24 is a diagram showing a projected basic area obtained by projecting a basic illumination area onto a predetermined virtual diffraction lens in the second embodiment.

FIG. 25A is a diagram showing a state in which the upper half of the divided projection basic region is divided into a large number of partial regions in the second embodiment, and FIG. 25B is a diagram corresponding to the large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 26 is a diagram showing how a basic optical element is formed by combining two divided basic optical elements in the second embodiment.

FIG. 27 is a diagram showing a state in which four types of basic optical elements in which the positions of the basic illumination areas are slightly different are arranged in the second embodiment.

FIG. 28 is a fly-eye integrator and a quad-pole aperture of a quad-pole aperture stop according to the third embodiment;
It is a figure which shows the relationship with the quadrupole illumination area | region which should be formed on the incident surface of an integrator.

FIG. 29 is a diagram showing a sub-projection basic area obtained by projecting the sub-basic illumination area onto a predetermined virtual diffraction lens in the third embodiment.

FIG. 30A is a diagram showing a state in which the upper half of the divided projection basic region is divided into a large number of partial regions in the third embodiment, and FIG. 30B corresponds to the large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 31 is a diagram showing how a sub basic optical element is formed by combining two divided basic optical elements and a basic optical element is formed by combining four sub basic optical elements in the third embodiment.

FIG. 32 is a diagram showing a state in which four types of basic optical elements in which the positions of the basic illumination areas are slightly different are arranged in the third embodiment.

FIG. 33 is a view showing a fly-eye integrator according to the fourth embodiment and a 2-pole aperture of the x-direction 2-pole aperture stop, and 2 to be formed on the incident surface of the fly-eye integrator.
It is a figure which shows the relationship with a polar illumination area.

FIG. 34 is a diagram showing a sub-projection basic area obtained by projecting the sub-basic illumination area on a predetermined virtual diffraction lens in the fourth embodiment.

FIG. 35 (a) is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions in the fourth embodiment, and FIG. 35 (b) corresponds to a large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 36 is a diagram showing a state in which a sub basic optical element is formed by combining two divided basic optical elements and a basic optical element is formed by combining two sub basic optical elements in the fourth embodiment.

FIG. 37 is a view showing a fly-eye integrator according to the fifth embodiment, a 2-pole aperture of a y-direction 2-pole aperture stop, and 2 to be formed on the incident surface of the fly-eye integrator;
It is a figure which shows the relationship with a polar illumination area.

FIG. 38 is a diagram showing a sub-projection basic area obtained by projecting the sub-basic illumination area on a predetermined virtual diffraction lens in the fifth embodiment.

FIG. 39A is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions in the fifth embodiment, and FIG. 39B corresponds to a large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 40 is a view showing a state in which a sub basic optical element is formed by combining two divided basic optical elements and a basic optical element is formed by combining two sub basic optical elements in the fifth embodiment.

FIG. 41 (a) is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions in the sixth embodiment, and FIG. 41 (b) corresponds to a large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 42A is a diagram showing a state in which the lower half divided projection basic region is divided into a large number of partial regions in the sixth embodiment, and FIG. 42B is a diagram corresponding to a large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

43 is a split basic optical element of FIG. 41 (b) and FIG.
It is a figure which shows a mode that a basic optical element is formed by composition with the division | segmentation basic optical element of (b).

FIG. 44 is a diagram showing a state in which four types of basic optical elements in which the positions of the basic illumination areas are slightly different are arranged in the sixth embodiment.

FIG. 45 is a diagram schematically showing a pattern formed on a mask used for manufacturing a diffractive optical element in the sixth embodiment by lithography.

FIG. 46 is a diagram schematically showing a configuration of a mask used for manufacturing a diffractive optical element in the sixth embodiment by lithography.

FIG. 47 is a diagram showing one diffractive optical element among a plurality of diffractive optical elements generated on a glass substrate using the mask of FIG. 46.

FIG. 48 is a diagram showing a relationship between an annular illumination region to be formed on the incident surface of the microlens array and a large number of lens elements in the seventh embodiment.

FIG. 49A is a diagram showing a state in which the upper half of the divided projection basic region is divided into a large number of partial regions in the seventh embodiment, and FIG. 49B is a diagram corresponding to the large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 50 is a diagram showing how a basic optical element is formed by combining two divided basic optical elements in the seventh embodiment.

FIG. 51A is a diagram showing a state in which the right-half divided projection basic region is divided into a large number of partial regions in the eighth embodiment, and FIG. 51B is a diagram showing a plurality of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 52 is a diagram showing how a basic optical element is formed by combining two divided basic optical elements in the eighth embodiment.

FIG. 53 is a diagram showing one diffractive optical element of a plurality of diffractive optical elements formed on a glass substrate in an eighth embodiment.

FIG. 54 is a diagram illustrating a first modified example according to the second embodiment.

FIG. 55 is a diagram illustrating a main part configuration of a second modified example according to the seventh embodiment and the eighth embodiment.

FIG. 56A is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions in the modified example of the sixth embodiment, and FIG. 56B is a diagram showing a plurality of divided partial regions. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.

FIG. 57 is a diagram showing a state of random arrangement in which indexes 1 to 548 are added to the 548 partial regions of FIG. 56A.

58 is a diagram showing a state of random arrangement in which indexes 1 to 548 are attached to the 548 partial optical elements of FIG. 56 (b).

FIG. 59 is a graph showing the illuminance distribution x in the modification of the sixth embodiment.
It is a figure explaining the method for improving the symmetry regarding a direction.

FIG. 60 is a diagram schematically showing a pattern formed on a mask used for manufacturing a diffractive optical element in a modified example of the sixth embodiment by lithography.

FIG. 61 is a diagram showing a diffractive optical element formed on a glass substrate using a mask manufactured based on the block pattern of FIG. 60.

FIG. 62 is a diagram showing a simulation result of a light intensity distribution when the diffractive optical element according to the modification of the sixth embodiment is applied to the exposure apparatus of FIG. 7.

63A is a diagram showing a state in which the upper half of the divided projection basic region is divided into a large number of partial regions in the modified example of the seventh embodiment, and FIG. 63B is a diagram showing a plurality of divided partial regions. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.

FIG. 64 (a) is a modification of the seventh embodiment shown in FIG.
FIG. 6 is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions assuming a virtual lens having a focal length different from 3, and (b) corresponds to a large number of divided partial regions. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements densely defined by the above.

FIG. 65 (a) is a modification of the seventh embodiment shown in FIG.
4 is a diagram showing a state in which the divided projection basic region obtained by rotating the divided projection basic region of 4 by 90 degrees in the xy plane is divided into a large number of partial regions, and FIG. It is a figure which shows the division | segmentation basic | foundation optical element obtained by rearranging many partial optical elements defined correspondingly densely.

FIG. 66 is a principle diagram illustrating a method of forming a basic optical element aggregate AP (j) by arranging basic optical elements P (j) and P (j) ′ _ 90 adjacent to each other in a modified example of the seventh embodiment. Is.

FIG. 67 is a principle diagram illustrating a method of forming a basic optical element assembly BP (j) by arranging basic optical elements P (j) and P (j) ′ _ 90 adjacent to each other in a modified example of the seventh embodiment. Is.

FIG. 68 is a diagram partially showing a pattern of the basic optical element P (1) in the modification of the seventh embodiment.

FIG. 69 is a diagram partially showing a pattern of a basic optical element P (2) according to a modification of the seventh embodiment.

FIG. 70 is a diagram schematically showing a pattern formed on a mask used for manufacturing a diffractive optical element according to a modification of the seventh embodiment by lithography.

71 is a diagram showing a diffractive optical element formed on a glass substrate using a mask manufactured based on the block pattern of FIG. 70.

72 is a diagram showing a simulation result of a light intensity distribution when the diffractive optical element according to the modification of the seventh embodiment is applied to the exposure apparatus of FIG. 7.

FIG. 73 (a) is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions in the ninth embodiment, and FIG. 73 (b) corresponds to a large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 74 is a diagram showing a simulation result of a light intensity distribution when the diffractive optical element of the ninth embodiment is applied to the exposure apparatus of FIG.

FIG. 75 (a) is a diagram showing a state in which the upper half divided projection basic region is divided into a large number of partial regions in the tenth embodiment, and FIG. 75 (b) corresponds to a large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 76 is a diagram showing a simulation result of a light intensity distribution when the diffractive optical element of the tenth embodiment is applied to the exposure apparatus of FIG.

77A is a diagram showing a state in which the upper half divided projection basic region is divided into a plurality of partial regions in the eleventh embodiment, and FIG. 77B is a diagram corresponding to a large number of divided partial regions. It is a figure which shows the division | segmentation basic optical element obtained by rearranging many defined partial optical elements densely.

FIG. 78 is a diagram showing a simulation result of a light intensity distribution when the diffractive optical element of the eleventh embodiment is applied to the exposure apparatus of FIG. 7.

79 (a) is a blazed phase Fresnel zone plate, (b) is an L-value phase Fresnel zone plate, FIG.
(C) is a cross-sectional view of the binary phase type Fresnel zone plate taken along a section including the center of rotation of the zone plate pattern.

FIG. 80 is a diagram schematically showing another configuration of an exposure apparatus to which the diffractive optical element according to each embodiment and each modification of the present invention can be applied.

FIG. 81 is a diagram schematically showing still another configuration of the exposure apparatus to which the diffractive optical element according to each embodiment and each modification of the present invention can be applied.

82 is a diagram schematically showing a configuration of a main part of the exposure apparatus of FIG. 81. FIG.

83 is a relationship between a fly-eye integrator having a configuration different from that of the fly-eye integrator in FIG. 7 and a ring-shaped aperture of a ring aperture stop, and a ring-shaped illumination area to be formed on the incident surface. It is a figure which shows the relationship between many lens elements.

FIG. 84 is a diagram showing how a semi-ring shaped divided basic illumination area is set in accordance with the necessary lens elements defined in FIG. 83.

85 is a state in which basic optical elements of different types having different internal arrangements generated based on the divided basic illumination areas defined in FIG. 84 are patterned so as to be included in the effective diameter of the diffractive optical element in the same number. FIG.

86 is a diagram schematically showing a projection basic region obtained based on a division pitch qy different from the projection basic region shown in FIG. 63.

[Explanation of symbols]

1 light source 2 beam shaping optics 3,7 Revolver 4 relay lens system 5 Vibration mirror 6 Fly-eye integrator 8 condenser lens system 9 Reticle (mask) 10 Projection optical system 11 wafers 21-26 Diffractive optical element 31-38 Aperture stop

─────────────────────────────────────────────────── ─── Continued Front Page (51) Int.Cl. ⁷ Identification Code FI Theme Coat (Reference) H01L 21/027 H01L 21/30 502P 515D

Claims (40)

[Claims] 1. A method of manufacturing a mask, which comprises a basic optical element composed of a plurality of partial optical elements and is used for manufacturing an optical element for converting an incident light beam into a light beam having a predetermined light beam cross section by lithography. And dividing the predetermined light beam cross section into a plurality of regions extending along the first direction to obtain a plurality of first light beam cross section partial regions, and the plurality of the plurality of regions obtained by the first dividing process. A total length calculation step of calculating a total length of the first light flux cross-section partial areas along the first direction, and a total length calculation obtained by the total length calculation step are equally divided to obtain a predetermined unit length. A unit length calculating step, dividing the plurality of first light flux cross-section partial regions based on the predetermined unit length obtained by the unit length calculating step,
A second division step of obtaining a plurality of second light flux cross-sectional partial areas, and a plurality of the predetermined optical elements using coordinate data on predetermined optical elements corresponding to the coordinates of the plurality of second light flux cross-sectional partial areas. A third dividing step of dividing into partial optical elements; a dense arrangement step of arranging the plurality of partial optical elements divided by the third dividing step densely to form the basic optical element; and a dense arrangement step And a drawing step of drawing a pattern for obtaining the obtained basic optical element by lithography on the mask. 2. A subdivision process for further dividing each of the plurality of first light flux cross-section partial regions to obtain a third light flux cross-section partial region, and a plurality of third light flux cross-section partial regions obtained by the subdivision process. The method for manufacturing a mask according to claim 1, further comprising a random arrangement step of randomly rearranging in the first light beam cross-section partial region. 3. The plurality of third light flux cross-section partial areas have a plurality of equal division partial areas and non-equal division partial areas, and each equal division partial area is set to a substantially square shape. The mask manufacturing method according to claim 1 or 2. 4. A regular arrangement step of regularly arranging the plurality of partial optical elements divided by the third dividing step, and the plurality of partial optical elements arranged regularly and densely by the regular arrangement step. 4. The mask manufacturing method according to claim 1, further comprising a random arrangement step of rearranging the masks at random. 5. The plurality of partial optical elements arranged by the regular arrangement step have at least one row, and the random arrangement step includes: in each row of the plurality of partial optical elements densely arranged. 5. The mask manufacturing method according to claim 4, further comprising the step of rearranging randomly. 6. The plurality of partial optical elements arranged by the regular arrangement step have at least two columns, and the random arrangement step randomly selects a row unit in the plurality of densely arranged partial optical elements. 6. The mask manufacturing method according to claim 4, further comprising the step of rearranging the mask. 7. The plurality of partial optical elements arranged by the regular arrangement step have at least two rows, and the random arrangement step comprises two rows of the plurality of partial optical elements densely arranged. 7. The method for manufacturing a mask according to claim 4, further comprising a step of exchanging partial optical elements having the same size. 8. A sectional dividing step of dividing the predetermined luminous flux section into a plurality of symmetrical luminous flux sections, and one section of the mutually symmetrical luminous flux sections obtained in the sectional dividing step, wherein: First division step, total length calculation step, unit length calculation step, second
A step of obtaining a first divided basic optical element by performing a dividing step, the third dividing step, and the dense arrangement step;
A step of obtaining one or more other divided basic optical elements based on the configuration of the first divided basic optical element and the symmetry of the plurality of light beam cross sections, and the first divided basic optical element and the other one 8. The mask manufacturing method according to claim 1, further comprising a synthesizing step of synthesizing a plurality of divided basic optical elements to form the basic optical element. 9. A shape setting step of setting a partial shape of the partial optical element in a state where the center of the predetermined light beam cross section and the center of the predetermined optical element in the third dividing step are decentered. The mask manufacturing method according to any one of claims 1 to 7, wherein: 10. The optical element comprises a first basic optical element and a second basic optical element, and in the shape setting step, when the first basic optical element is obtained, A first shape setting step of setting a center and a center of the predetermined optical element in the third division step to a first eccentric state; and a predetermined cross section of the light flux when obtaining the second basic optical element. 10. The mask manufacturing method according to claim 9, further comprising a second shape setting step of setting a center and a center of the predetermined optical element in the third dividing step to a second eccentric state. 11. The first eccentric state in the first shape setting step and the second eccentric state in the second shape setting step.
11. The mask manufacturing method according to claim 10, wherein the eccentric state is symmetrical with each other. 12. The optical element includes a first basic optical element and a second basic optical element, and the second basic optical element includes the plurality of partial optical elements in the first basic optical element. 10. The mask manufacturing method according to claim 9, wherein the partial shape data of the element is generated by performing coordinate conversion according to a predetermined rule. 13. The predetermined optical element is a binary phase type Fresnel zone plate, and the basic optical element is configured based on half of the predetermined light beam cross section. The method for manufacturing a mask according to any one of items. 14. The first basic optical element, wherein the basic optical element is configured based on a first light flux cross section corresponding to half of the predetermined light flux cross section, and the first light flux cross section is 90 degrees around the optical axis. 15. The mask manufacturing method according to claim 13, further comprising a second basic optical element configured based on a second light beam cross section obtained by rotating the second light beam. 15. A method of manufacturing a mask, which is provided with a basic optical element composed of a plurality of partial optical elements and is used for manufacturing an optical element for converting an incident light beam into a light beam having a predetermined light beam cross section by lithography. And a first dividing step of dividing the predetermined light beam cross section into a plurality of areas to obtain a plurality of light beam cross section partial areas, and coordinate data on a predetermined optical element corresponding to coordinates of the plurality of light beam cross section partial areas. A second dividing step of dividing the predetermined optical element into a plurality of partial optical elements by using, and the plurality of partial optical elements divided by the second dividing step are densely arranged to form the basic optical element. And a drawing step of drawing on the mask a pattern for obtaining the basic optical element obtained by the dense arrangement step by lithography, The dividing step includes a step of determining coordinate data of each of the plurality of partial optical elements on the predetermined optical element, and a step of associating the shape data of each of the plurality of partial optical elements with the coordinate data. The step of obtaining the coordinate conversion data of the partial optical elements based on the coordinate data of the plurality of partial optical elements that are densely arranged and the coordinate data in the second dividing step, A step of performing coordinate transformation of the shape data on the basis of the coordinate transformation data. 16. The second dividing step includes a step of expressing the shape data of each of the plurality of partial optical elements in a predetermined pattern, and in the step of converting the coordinates of the shape data, the predetermined pattern is changed. The mask manufacturing method according to claim 15, wherein the coordinates are converted. 17. The second dividing step comprises a mesh dividing step of dividing the plurality of partial optical elements into a large number of cells, and a plurality of cells obtained by the mesh dividing step and the plurality of partial optical elements. Based on each of the predetermined pattern, including a determining step of determining a cell to form a part of the pattern and a cell not forming a part of the pattern, in the step of coordinate conversion of the shape data, The mask manufacturing method according to claim 16, wherein the cell information obtained in the determining step is subjected to coordinate conversion. 18. A mask manufactured by the manufacturing method according to any one of claims 1 to 17. 19. An optical element manufactured by using the mask according to claim 18. 20. The optical element according to claim 19, wherein the optical element is a diffractive optical element, a refractive optical element, or a reflective optical element. 21. A mask used for lithographically manufacturing an optical element comprising a basic optical element composed of a plurality of partial optical elements, the optical element converting an incident light beam into a light beam having a predetermined light beam cross section. A mask having an element pattern and a control regular pattern for controlling at least one of a line width and a depth. 22. A method of manufacturing an optical element, comprising a basic optical element composed of a plurality of partial optical elements, for converting an incident light beam into a light beam having a predetermined light beam cross section, wherein the mask on the mask according to claim 21. A method of manufacturing an optical element, comprising a pattern transfer step of transferring a pattern onto a substrate coated with a photosensitive material. 23. The optical element manufacturing method according to claim 22, wherein in the pattern transfer step, the line width of the transferred pattern is controlled using the control rule pattern. 24. The method further comprises a developing step of developing the substrate that has undergone the pattern transferring step, and an etching step of etching the substrate that has undergone the developing step, wherein the control step transferred to the substrate is performed in the etching step. 24. The method of manufacturing an optical element according to claim 22, wherein the etching is controlled based on the depth of the rule pattern for use. 25. The mask has a cutting guide pattern, and the pattern transfer step includes a step of exposing the cutting guide pattern to the substrate, wherein the cutting guide pattern is exposed to the substrate. The optical element manufacturing method according to claim 22, further comprising a cutting step of cutting the substrate along the line. 26. An optical element manufactured by the manufacturing method according to claim 22. 27. A basic optical element comprising a plurality of partial optical elements, wherein the plurality of partial optical elements convert an incident light beam into a plurality of partial light beams, and the basic optical element converts the incident light beam into the plurality of partial light beams. An optical element for converting into a light flux having a predetermined light flux cross section as a sum of light fluxes, wherein the basic optical element is set to correspond to a plurality of partial regions obtained by dividing the predetermined light flux cross section. An optical element characterized in that the partial optical elements of are densely and randomly arranged. 28. The method according to claim 19, wherein the plurality of partial optical elements have a diffraction pattern, and the line width of the diffraction pattern is larger than 0.1 μm and smaller than 1000 μm. Optical element. 29. An illumination optical device for illuminating a mask on which a predetermined pattern is formed, comprising: a light source for supplying an illumination light beam; and the optical element according to claim 19, 20, 26, 27 or 28. A light flux conversion unit for converting an illumination light flux from the light source into a light flux having a predetermined light flux cross section and guiding it to a predetermined surface, and a surface having a predetermined shape based on the light flux from the light flux conversion unit having the predetermined light flux cross section. An illumination optical apparatus comprising: an optical integrator for forming a light source; and a light guide optical system for guiding a light beam from the optical integrator to the mask. 30. The optical element in the light flux converting section,
30. The illumination optical device according to claim 29, comprising a plurality of the basic optical elements. 31. The plurality of basic optical elements include a first basic optical element and a second basic optical element, and each of the first basic optical element and the second basic optical element is the predetermined optical element. A plurality of partial optical elements set corresponding to a plurality of partial areas obtained by dividing the luminous flux cross section of the optical element are densely and randomly arranged, and the plurality of partial optical elements in the first basic optical element are arranged. 31. The illumination optical apparatus according to claim 30, wherein the random arrangement of the partial optical elements of <1> and the random arrangement of the plurality of partial optical elements of the second basic optical element are different from each other. 32. The light flux conversion section includes a plurality of optical elements switchable with respect to an illumination optical path in order to change a light intensity distribution on the predetermined surface. 32. The illumination optical device according to any one of 31. 33. The optical integrator comprises:
The optical element has a wavefront division type optical integrator having a plurality of optical elements arranged two-dimensionally, and a division pitch for dividing the predetermined light beam cross section into the plurality of first light beam cross section partial regions is an arrangement pitch of the optical elements. 33. The illumination optical device according to claim 29, wherein the illumination optical device is set based on 34. In order to set at least one of an illuminance distribution on a conjugate plane of the mask and an illuminance distribution on an illumination pupil plane to a desired state, the basic optical element includes a plurality of densely arranged partial optical elements. The illumination optical system according to any one of claims 29 to 33, characterized in that the array of elements is set as an initial value and the array of the partial optical elements is optimized by using a predetermined algorithm. apparatus. 35. In order to set at least one of an illuminance distribution on a conjugate plane of the mask and an illuminance distribution on an illumination pupil plane to a desired state, the basic optical element includes a plurality of densely arranged partial optics. Set by initializing the array of elements and the internal pattern, further subdividing each partial optical element, and optimizing the array of the partial optical elements and the array of the subdivided internal pattern using a predetermined algorithm. Claim 2 characterized in that
The illumination optical device according to any one of 9 to 34. 36. An illumination optical device according to claim 29, and a projection optical system for projecting the pattern image of the mask onto a substrate coated with a photosensitive material. An exposure apparatus characterized by the above. 37. An illuminating step of illuminating the mask with the illumination optical device according to claim 29, and a projecting step of projecting a pattern image of the mask onto a substrate coated with a photosensitive material. And an exposure method comprising: 38. In a mask in which a basic pattern to be exposed is drawn on a substrate coated with a photosensitive material, a regular pattern for controlling the line width of the pattern exposed on the substrate and the cutting of the substrate A mask characterized in that a cutting guide pattern for guiding is drawn in parallel. 39. The mask according to claim 38, wherein the regular pattern is a line and space pattern or a dot pattern. 40. The mask is used to etch the substrate exposed with the basic pattern to form a desired pattern, and the regular pattern is used to control a depth of the etching. 40. The mask according to claim 38 or 39.