Pattern exposure apparatus, component manufacturing method, and exposure apparatus

Abstract

A pattern exposure apparatus includes: a spatial light modulator having a plurality of micromirrors selectively driven based on drawing data; an illumination unit that illuminates the spatial light modulator with illumination light at a predetermined incident angle; and a projection unit that projects reflected light from the micromirrors in a selected open state of the spatial light modulator onto a substrate; wherein the pattern exposure apparatus projects and exposes a pattern corresponding to the drawing data onto the substrate, and the illumination unit includes: a focusing optical component that focuses light from a surface light source of a predetermined shape that serves as the source of illumination light onto the spatial light modulator at an oblique angle, and is arranged along an optical axis oblique to the optical axis of the projection unit at an incident angle, for optically conjugating the surface light source and the pupil of the projection unit; and a correction optical component that corrects the shape deformation of the outline of the surface light source in such a way that the outline of the image of the surface light source formed on the pupil of the projection unit by the reflected light from the micromirrors in the open state of the spatial light modulator is deformed into an elliptical shape according to the incident angle.

Description

[Technical Field]

[0001] This invention relates to a pattern exposure apparatus for exposing patterns for electronic components, a method for manufacturing electronic components using such a pattern exposure apparatus, and an exposure apparatus. This application claims priority based on Japanese Patent Application No. 2021-111450, filed on July 5, 2021, the contents of which are incorporated herein by reference. [Previous Technology]

[0002] Previously, in the lithography process for manufacturing electronic components (micro-components) such as liquid crystal or organic EL (Electroluminescence) display panels and semiconductor devices (integrated circuits, etc.), projection exposure apparatuses using a step-and-repeat method (so-called steppers) or projection exposure apparatuses using a step-and-scan method (so-called scanning steppers (also called scanners)) were used. These exposure apparatuses project and expose a mask pattern for electronic components onto a photosensitive layer coated on the surface of the substrate to be exposed (hereinafter also simply referred to as the substrate), such as a glass substrate, semiconductor wafer, printed wiring substrate, or resin film.

[0003] Fabricating a mask substrate that permanently forms the mask pattern requires time and expense. Therefore, an exposure apparatus is known that uses a spatial light modulation device (variable mask pattern generator) such as a digital mirror element (DMD) with a plurality of micromirrors arranged in a regular pattern (see, for example, Patent Document 1) instead of a mask substrate. In the exposure apparatus disclosed in Patent Document 1, light from a semiconductor laser light source 3 with a wavelength of 405 nm or 365 nm is obliquely irradiated at an incident angle of 22° to 26° by an irradiation optical system 6 onto the digital mirror element (DMD) which serves as a spatial light modulator 4. The reflected light from the multiple pixel mirrors of the spatial light modulator 4 (DMD) in the open state is projected onto the exposure area of ​​the object W by a projection optical system 5.

[0004] In the case of Patent Document 1, the tilt angle of the pixel mirror (micromirror) of the DMD is set to half an angle of 22° to 26° of the incident angle of the illumination light. A plurality of pixel mirrors (micromirrors) are arranged in a matrix at a certain spacing, thus also functioning as an optical diffraction grating (blazed diffraction grating). Especially in the case of projecting and exposing fine patterns for electronic components, when the illumination light directed towards the DMD is tilted, the DMD's function as a diffraction grating (the direction or intensity distribution of the diffracted light) sometimes leads to a deterioration in the image quality of the pattern. [Prior Art Documents] [Patent Documents]

[0005] [Patent Document 1] International Publication No. 2018 / 088550 [Summary of the Invention]

[0006] According to a first aspect of the present invention, a pattern exposure apparatus is provided, comprising: a spatial light modulation device having a plurality of micromirrors selectively driven based on drawing data; an illumination unit illuminating the spatial light modulation device with illumination light at a predetermined incident angle; and a projection unit projecting reflected light from the micromirrors in a selected on state of the spatial light modulation device onto a substrate; wherein the pattern exposure apparatus projects and exposes a pattern corresponding to the drawing data onto the substrate, and the illumination unit comprises: a focusing optical component that directs light from the micromirrors in a selected on state onto the substrate. The light from a surface light source of a predetermined shape, which is the source of the illumination light, is focused and obliquely irradiated onto the aforementioned spatial light modulation device. The surface light source is arranged along an optical axis that is oblique to the optical axis of the aforementioned projection unit at the aforementioned angle of incidence, so as to make the surface light source optically conjugate with the pupil of the aforementioned projection unit. The optical component is used to correct the shape deformation of the outline of the surface light source in a manner that corrects the situation where the outline of the image of the surface light source formed on the pupil of the aforementioned projection unit by the reflected light from the micromirror of the aforementioned spatial light modulation device in the aforementioned open state is deformed into an elliptical shape according to the aforementioned angle of incidence.

[0007] According to a second aspect of the present invention, a method for manufacturing an electronic component is provided, which uses illumination light from an illumination unit to illuminate a spatial light modulator having a plurality of micromirrors having selectively driven drawing data based on a pattern for an electronic component, and projects reflected light from the selected on state of the micromirrors of the spatial light modulator onto a substrate via a projection unit, thereby forming a pattern of the electronic component on the substrate. The method for manufacturing the electronic component includes the following stages: focusing light from a surface light source of a predetermined shape generated in the illumination unit as the illumination light and obliquely illuminating the spatial light modulator at a predetermined incident angle; and deforming the shape of the outline of the surface light source in a manner that corrects the situation where the outline of the image of the surface light source formed on the pupil of the projection unit by the reflected light from the on state of the micromirrors of the spatial light modulator is deformed into an elliptical shape according to the incident angle.

[0008] According to a third aspect of the present invention, an exposure apparatus is provided, comprising: a spatial light modulation device having a plurality of micromirrors; an illumination unit that illuminates the spatial light modulation device with illumination light; and a projection unit that projects reflected light from the micromirrors in the open state of the spatial light modulation device; wherein the illumination unit comprises: a cylindrical lens; and a focusing optical component that focuses the light passing through the cylindrical lens and illuminates the spatial light modulation device.

Implementation Method

[0010] A preferred embodiment of the pattern exposure apparatus (pattern forming apparatus) of the present invention will be disclosed, and a detailed description will be provided below with reference to the drawings. Furthermore, the present invention is not limited to these embodiments, but also includes embodiments with various modifications or improvements. That is, the constituent elements described below include constituent elements that can be easily conceived by the industry, substantially the same constituent elements, and constituent elements that can be appropriately combined. Moreover, various omissions, substitutions, or changes to constituent elements can be made without departing from the spirit of the present invention. Furthermore, throughout the drawings and the following detailed description, the same reference numerals are used for components or constituent elements that achieve the same or identical functions.

[0011] [Overall Structure of Pattern Exposure Apparatus] FIG1 is a perspective view showing the outline of the external structure of the pattern exposure apparatus (hereinafter also referred to as the exposure apparatus) EX of this embodiment. The exposure apparatus EX is a device for projecting an image onto an exposed substrate using exposure light, and the intensity distribution of the exposure light in the space is dynamically modulated by a spatial light modulation device (digital mirror element: DMD). In a specific embodiment, the exposure apparatus EX is a projection exposure apparatus (scanner) using a step-scan method, and a rectangular (square) glass substrate used in display devices (flat panel displays) is used as the exposure object. The glass substrate is a flat panel display substrate P with at least one side length or diagonal length of 500 mm or more and a thickness of 1 mm or less. The exposure apparatus EX exposes the projected image of the pattern produced by the DMD to a photosensitive layer (photoresist) formed on the surface of the substrate P with a certain thickness. After exposure, the substrate P, which is removed from the exposure device EX, is sent to a predetermined process step (film formation step, etching step, plating step, etc.) after the development step.

[0012] The exposure apparatus EX includes a platform assembly comprising: a platform 2 mounted on active vibration damping units 1a, 1b, 1c, and 1d (1d not shown); a plate 3 mounted on the platform 2; an XY platform 4A movable in two dimensions on the plate 3; a substrate holder 4B held on the XY platform 4A in a plane; and laser length measuring interferometers (hereinafter also referred to as interferometers) IFX, IFY1 to IFY4, which measure the two-dimensional movement position of the substrate holder 4B (substrate P). Such a platform assembly is disclosed, for example, in U.S. Patent Publication No. 2010 / 0018950 and U.S. Patent Publication No. 2012 / 0057140.

[0013] In Figure 1, the XY plane of the orthogonal coordinate system XYZ is set to be parallel to the flat surface of the plate 3 of the platform device, and the XY platform 4A is set to be able to translate within the XY plane. Furthermore, in this embodiment, the direction parallel to the X-axis of the coordinate system XYZ is set as the scanning movement direction of the substrate P (XY platform 4A) during scanning exposure. The movement position of the substrate P in the X-axis direction is measured sequentially by interferometers IFX, and the movement position in the Y-axis direction is measured sequentially by at least one (preferably two) of four interferometers IFY1 to IFY4. The substrate holder 4B is configured to be able to move slightly relative to the XY platform 4A in the Z-axis direction perpendicular to the XY plane, and to be able to tilt slightly relative to the XY plane in any direction, actively performing focus adjustment and leveling (parallelism) adjustment between the surface of the substrate P and the imaging surface of the projected pattern. Furthermore, the substrate holder 4B is configured to actively adjust the tilt of the substrate P in the XY plane by means of being able to rotate slightly (θz rotation) around an axis parallel to the Z axis.

[0014] The exposure apparatus EX further comprises: an optical plate 5, which holds a plurality of exposure (drawing) module groups MU(A), MU(B), and MU(C); and main columns 6a, 6b, 6c, and 6d (6d not shown), which support the optical plate 5 from the base 2. Each of the plurality of exposure module groups MU(A), MU(B), and MU(C) comprises: an illumination unit ILU, which is mounted on the +Z direction side of the optical plate 5 and receives illumination light from the fiber optic unit FBU; and a projection unit PLU, which is mounted on the -Z direction side of the optical plate 5 and has an optical axis parallel to the Z-axis. Furthermore, each of the exposure module groups MU(A), MU(B), and MU(C) comprises: a digital mirror element (DMD) 10 serving as an optical modulation unit, which reflects the illumination light from the illumination unit ILU in the -Z direction and receives it from the projection unit PLU. The detailed configuration of the exposure module group obtained from the illumination unit ILU, DMD 10, and projection unit PLU will be described later.

[0015] On the -Z direction side of the optical plate 5 of the exposure apparatus EX, a plurality of alignment systems (microscopes) ALGs are mounted, which detect alignment marks formed on the substrate P at predetermined positions. In order to confirm (correct) the relative positional relationship of the detection fields of the various alignment systems ALGs in the XY plane, to confirm (correct) the baseline error between the projection positions of the pattern images projected by the projection units PLU of each exposure module group MU(A), MU(B), and MU(C) and the positions of the detection fields of each alignment system ALG, or to confirm the position or image quality of the pattern images projected by the projection units PLU, a correction reference part CU is provided at the end of the substrate holder 4B in the -X direction. Furthermore, although a portion is not shown in FIG1, in this embodiment, as an example, each of the exposure module groups MU(A), MU(B), and MU(C) has 9 modules arranged at certain intervals in the Y direction, but the number of modules may be less than 9 or more than 9.

[0016] Figure 2 is a diagram showing an example of the arrangement of the projection area IAn of the digital mirror element (DMD) 10 projected onto the substrate P by the projection units PLU of each of the exposure module groups MU(A), MU(B), and MU(C). The orthogonal coordinate system XYZ is set to be the same as in Figure 1. In this embodiment, the first row of exposure module groups MU(A), the second row of exposure module groups MU(B), and the third row of exposure module groups MU(C), which are arranged away from each other in the X direction, are each composed of 9 modules arranged along the Y direction. The exposure module group MU(A) is composed of 9 modules MU1 to MU9 arranged in the +Y direction, the exposure module group MU(B) is composed of 9 modules MU10 to MU18 arranged in the -Y direction, and the exposure module group MU(C) is composed of 9 modules MU19 to MU27 arranged in the +Y direction. Modules MU1 to MU27 are all identical in configuration. When exposure module group MU(A) and exposure module group MU(B) are set to face each other in the X direction, exposure module group MU(B) and exposure module group MU(C) become back-to-back in the X direction.

[0017] In Figure 2, as an example, the projection areas IA1, IA2, IA3, ..., IA27 (where n is set to 1 to 27, or sometimes represented as IAn) of modules MU1 to MU27 are rectangles with an aspect ratio of approximately 1:2 and extending in the Y direction. In this embodiment, as the substrate P moves in the +X direction, the -Y direction ends of the first column projection areas IA1 to IA9 and the +Y direction ends of the second column projection areas IA10 to IA18 are successively exposed. In addition, the areas on the substrate P that are not exposed by the first and second column projection areas IA1 to IA18 are successively exposed by the third column projection areas IA19 to IA27. The center points of the first column of projection areas IA1 to IA9 are located on line k1, which is parallel to the Y-axis. The center points of the second column of projection areas IA10 to IA18 are located on line k2, which is parallel to the Y-axis. The center points of the third column of projection areas IA19 to IA27 are located on line k3, which is parallel to the Y-axis. The X-direction interval between lines k1 and k2 is set to distance XL1, and the X-direction interval between lines k2 and k3 is set to distance XL2.

[0018] Here, when the connection between the end of the projection area IA9 in the -Y direction and the end of the projection area IA10 in the +Y direction is set as OLa, the connection between the end of the projection area IA10 in the -Y direction and the end of the projection area IA27 in the +Y direction is set as OLb, and the connection between the end of the projection area IA8 in the +Y direction and the end of the projection area IA27 in the -Y direction is set as OLc, the state of continuous exposure is explained with reference to FIG3. In FIG3, the orthogonal coordinate system XYZ is set as in FIG1 and FIG2, and the coordinate system X'Y' in the projection areas IA8, IA9, IA10, IA27 (and all other projection areas IAn) is set with an inclination angle θk relative to the X-axis and Y-axis (lines k1 to k3) of the orthogonal coordinate system XYZ. That is, by arranging the multiple micromirrors of DMD10 in a two-dimensional coordinate system X'Y, the entire DMD10 is tilted at an angle θk in the XY plane.

[0019] The circular area in Figure 3 containing the projection areas IA8, IA9, IA10, IA27 (and all other projection areas IAn) represents the circular image field PLf' of the projection unit PLU. In the connecting part Ola, the projection image of the micromirrors arranged at an angle (θk) at the end of the -Y' direction of projection area IA9 is set to overlap with the projection image of the micromirrors arranged at an angle (θk) at the end of the +Y' direction of projection area IA10. In the connecting part OLb, the projection image of the micromirrors arranged at an angle (θk) at the end of the -Y' direction of projection area IA10 is set to overlap with the projection image of the micromirrors arranged at an angle (θk) at the end of the +Y' direction of projection area IA27. Similarly, in the connecting part OLc, the projection image of the micromirrors arranged at an angle (θk) at the end of the projection area IA8 in the +Y' direction is set to overlap with the projection image of the micromirrors arranged at an angle (θk) at the end of the projection area IA27 in the -Y' direction.

[0020] [Composition of the Illumination Unit] Figure 4 is an optical configuration diagram showing the specific composition of module MU18 in exposure module group MU(B) and module MU19 in exposure module group MU(C) as shown in Figures 1 and 2, viewed in the XZ plane. The orthogonal coordinate system XYZ in Figure 4 is set to be the same as the orthogonal coordinate system XYZ in Figures 1 to 3. Furthermore, as shown by the configuration of each module in the XY plane in Figure 2, module MU18 is offset from module MU19 in the +Y direction at a certain interval and is arranged back-to-back with each other. The optical components in module MU18 and the optical components in module MU19 are made of the same material and are constructed in the same manner. Therefore, the optical composition of module MU18 will be described in detail here. Furthermore, the fiber unit FBU shown in Figure 1 corresponds to each of the 27 modules MU1 to MU27 shown in Figure 2 and is composed of 27 fiber bundles FB1 to FB27.

[0021] The illumination unit ILU of module MU18 is composed of the following parts: mirror 100, which reflects the illumination light ILm traveling in the -Z direction from the emitting end of fiber bundle FB18; mirror 102, which reflects the illumination light ILm from mirror 100 in the -Z direction; input lens system 104, which functions as a collimating lens; illuminance adjustment filter 106; optical integrator 108, which includes a micro compound eye (MFE) lens or field lens, etc.; condenser lens system 110; and tilting mirror 112, which reflects the illumination light ILm from condenser lens system 110 towards DMD10. Mirror 102, input lens system 104, optical integrator 108, condenser lens system 110 and tilting mirror 112 are arranged along the optical axis AXc parallel to the Z-axis.

[0022] The fiber bundle FB18 is formed by combining one or more fiber lines. The illumination light ILm emanating from the emitting end of the fiber bundle FB18 (each fiber line) is set to a numerical aperture (NA, also known as the diffusion angle) that is incident on the input lens system 104 at the rear and is not repelled. The position of the front focal point of the input lens system 104 is designed to be the same as the position of the emitting end of the fiber bundle FB18. Furthermore, the position of the rear focal point of the input lens system 104 is set in such a way that the illumination light ILm from one or more point sources formed at the emitting end of the fiber bundle FB18 overlaps on the incident surface side of the MFE lens 108A of the optical integrator 108. Therefore, the incident surface of the MFE lens 108A is Koehler illuminated by the illumination light ILm from the emitting end of the fiber bundle FB18. Furthermore, in the initial state, assume that the geometric center point of the output end of the fiber bundle FB18 in the XY plane is located on the optical axis AXc, and the principal ray (center line) of the illumination light ILm from the point source at the output end of the fiber is parallel (or coaxial) with the optical axis AXc.

[0023] The illumination light ILm from the input lens system 104 is attenuated by the illuminance adjustment filter 106 at an arbitrary value within the range of 0% to 90%, and then enters the condenser lens system 110 through the optical integrator 108 (MFE lens 108A, field lens, etc.). The MFE lens 108A is composed of multiple rectangular microlenses of tens of μm square arranged in two dimensions, and its overall shape is set in a manner that is approximately similar to the overall shape of the mirror surface of the DMD 10 in the XY plane (with an aspect ratio of approximately 1:2). Furthermore, the position of the front focal point of the condenser lens system 110 is set in a manner that is approximately the same as the position of the exit surface of the MFE lens 108A. Therefore, the illumination light from the point source on each of the multiple microlenses formed on the MFE lens 108A is transformed into approximately parallel beams by the condenser lens system 110, and after being reflected by the tilting mirror 112, they overlap on the DMD 10 to form a uniform illuminance distribution.

[0024] A surface light source is generated on the emission surface of the MFE lens 108A, which is formed by a dense two-dimensional arrangement of a plurality of point light sources (focusing points), thus functioning as a surface light source component. Such an MFE lens 108A can also be configured as disclosed in Japanese Patent Application Publication No. 2004-045885, for example, with two cylindrical microcompound eye lens devices arranged in the optical axis direction. The cylindrical microcompound eye lens devices are formed by arranging a plurality of cylindrical lenses on the incident surface side and the emission surface side of the illumination light, respectively.

[0025] Within the module MU18 shown in Figure 4, the optical axis AXc, parallel to the Z-axis, of the focusing lens system 110 is bent by the tilting mirror 112 to reach the DMD10, and the optical axis between the tilting mirror 112 and the DMD10 is set as the optical axis AXb. In this embodiment, the neutral plane of the center point of each of the plurality of micromirrors including the DMD10 is set to be parallel to the XY plane. Therefore, the angle between the normal of the neutral plane (parallel to the Z-axis) and the optical axis AXb becomes the incident angle θα of the illumination light ILm on the DMD10. The DMD10 is mounted on the lower side of the fixing part 10M, which is fixed to the support column of the illumination unit ILU. In the fixing part 10M, in order to finely adjust the position or posture of the DMD10, a micro-motion platform combining a parallel linkage mechanism and a retractable piezoelectric device, as disclosed in International Patent Publication No. 2006 / 120927, is provided.

[0026] Illumination light ILm from the micromirror in the On state of the DMD10 is reflected in the X direction of the XZ plane, directed toward the projection unit PLU. On the other hand, illumination light ILm from the micromirror in the Off state of the DMD10 is reflected in the Y direction of the YZ plane, not toward the projection unit PLU. Details will be described later. In this embodiment, the DMD10 is configured to switch between the On and Off states by tilting the micromirror in the roll and pitch directions.

[0027] A movable shutter 114 is detachably provided in the optical path between the DMD10 and the projection unit PLU. This movable shutter 114 is used to block reflected light from the DMD10 during non-exposure periods. As illustrated on the module MU19 side, the movable shutter 114 rotates to an angle position that retracts from the optical path during exposure periods, and rotates to an angle position that is tilted into the optical path during non-exposure periods, as illustrated on the module MU18 side. A reflective surface is formed on the DMD10 side of the movable shutter 114, where reflected light from the DMD10 irradiates the light absorber 115. The light absorber 115 does not reflect light energy in the ultraviolet wavelength range (wavelengths below 400 nm) again, but absorbs it and converts it into heat energy. Therefore, a heat dissipation mechanism (heat sink or cooling mechanism) is also provided on the light absorber 115. Furthermore, although not shown in Figure 4, the reflected light from the micromirror of the DMD10, which is in the Off state, during the exposure period is absorbed by the same light absorber (not shown in Figure 4) located in the Y direction (orthogonal to the plane of the paper in Figure 4) relative to the optical path between the DMD10 and the projection unit PLU.

[0028] [Structure of Projection Unit] The projection unit PLU mounted on the lower side of the optical plate 5 is configured as a telecentric imaging projection lens system. This telecentric imaging projection lens system is composed of a first lens group 116 and a second lens group 118 arranged along the optical axis AXa, which is parallel to the Z-axis. The first lens group 116 and the second lens group 118 are configured to be translated relative to the support column fixed on the lower side of the optical plate 5 by micro-actuators in the direction along the Z-axis (optical axis AXa). The projection magnification Mp of the imaging projection lens system obtained by the first lens group 116 and the second lens group 118 is determined by the relationship between the arrangement spacing Pd of the micromirrors on the DMD10 and the minimum linewidth (minimum pixel size) Pg of the pattern in the projection area IAn (n=1~27) projected on the substrate P.

[0029] As an example, when the required minimum linewidth (minimum pixel size) Pg is 1 μm and the micromirror spacing Pd is 5.4 μm, the tilt angle θk of the projection area IAn (DMD10) in the XY plane as described in Figure 3 above is also considered, and the projection magnification Mp is set to approximately 1 / 6. The imaging projection lens system obtained by the lens group 116, 118 causes the reduced image of the entire mirror surface of DMD10 to be inverted / reversed and imaged on the projection area IA18 (IAn) on the substrate P.

[0030] The first lens group 116 of the projection unit PLU can be slightly moved in the optical axis AXa direction by an actuator to perform fine adjustment of the projection magnification Mp (approximately ± tens of ppm). The second lens group 118 can be slightly moved in the optical axis AXa direction by an actuator to perform high-speed focus adjustment. Furthermore, in order to measure the positional change of the substrate P surface in the Z-axis direction with sub-micron precision, a plurality of oblique incident light type focus sensors 120 are provided on the underside of the optical plate 5. The plurality of focus sensors 120 measure the overall positional change of the substrate P in the Z-axis direction, the positional change of a portion of the substrate P corresponding to each projection area IAn (n = 1 to 27) in the Z-axis direction, or the local tilting change of the substrate P, etc.

[0031] As illustrated in Figure 3 above, the illumination unit ILU and projection unit PLU of the above type require the projection area IAn to be tilted at an angle θk in the XY plane. Therefore, the DMD10 and illumination unit PLU (at least the optical path portion of mirrors 102 to 112 along the optical axis AXc) in Figure 4 are configured in a manner where the whole is tilted at an angle θk in the XY plane.

[0032] Figure 5 is a schematic diagram showing the tilt angle θk of DMD10 and illumination unit PLU in the XY plane. In Figure 5, the orthogonal coordinate system XYZ is the same as the coordinate system XYZ of each of Figures 1 to 4 above, and the arrangement coordinate system X'Y' of the micromirrors Ms of DMD10 is the same as the coordinate system X'Y' shown in Figure 3. The circle containing DMD10 is the image field PLf of the object side of projection unit PLU, and the optical axis Axa is located at its center. On the other hand, the optical axis AXb obtained by bending the optical axis AXc of the condenser lens system 110 of illumination unit ILU by tilting mirror 112 is arranged in a way that tilts from the line Lu parallel to the X-axis at an angle θk when viewed in the XY plane.

[0033] [Imaging Optical Path Obtained by DMD] Next, referring to FIG6, the imaging state of the micromirror Ms of DMD10 obtained by the projection unit PLU (imaging projection lens system) will be described in detail. The orthogonal coordinate system X'Y'Z of FIG6 is the same as the coordinate system X'Y'Z shown in FIG3 and FIG5 above. FIG6 illustrates the optical path from the condenser lens system 110 of the illumination unit ILU to the substrate P. The illumination light ILm from the condenser lens system 110 travels along the optical axis AXc, is totally internally reflected by the tilting mirror 112, and reaches the mirror surface of DMD10 along the optical axis AXb. Here, the micromirror Ms located at the center of DMD10 is designated as Msc, and the micromirror Ms located at the periphery is designated as Msa. These micromirrors Msc and Msa are set to the on state.

[0034] Regarding the tilt angle of the micromirror Ms in the On state, if 17.5° is set as the standard value relative to the X'Y' plane (XY plane), then in order to make the principal rays of the reflected light Sc and Sa from the micromirrors Msc and Msa parallel to the optical axis AXa of the projection unit PLU, the incident angle (the angle between the optical axis AXb and the optical axis AXa) of the illumination light ILm illuminating the DMD10 is set to 35.0°. Therefore, in this case, the reflecting surface of the tilt mirror 112 is also tilted by 17.5° (=θα / 2) relative to the X'Y' plane (XY plane). The principal ray Lc of the reflected light Sc from the micromirror Msc is coaxial with the optical axis AXa, and the principal ray La of the reflected light Sa from the micromirror Msa is parallel to the optical axis AXa. The reflected light Sc and Sa are incident on the projection unit PLU with a predetermined numerical aperture (NA).

[0035] By means of reflected light Sc, the reduced image ic of the micromirror Msc, which is reduced by the projection magnification Mp of the projection unit PLU, is imaged on the substrate P in a telecentric state at a position on the optical axis AXa. Similarly, by means of reflected light Sa, the reduced image ia of the micromirror Msa, which is reduced by the projection magnification Mp of the projection unit PLU, is imaged on the substrate P in a telecentric state at a position away from the reduced image ic in the +X' direction. As an example, the first lens system 116 of the projection unit PLU is composed of three lens groups G1, G2, and G3, and the second lens system 118 is composed of two lens groups G4 and G5. An exit pupil (also simply called a pupil) Ep is provided between the first lens system 116 and the second lens system 118. At the position of the pupil Ep, a light source image of illumination light ILm (a collection of multiple point light sources formed on the exit surface side of the MFE lens 108A) is formed, which constitutes the Köhler illumination. The pupil Ep, also known as the opening of the projection unit PLU, is a key factor in determining the resolution of the projection unit PLU.

[0036] The orthogonal reflected light system from the micromirror Ms in the on state of DMD10 is set to pass through without being blocked by the maximum aperture (diameter) of the pupil Ep. The numerical aperture NAi on the image side (substrate P side) in the formula R = k1·(λ / NAi) which represents the resolution R is determined by the distance between the maximum aperture of the pupil Ep and the focal point on the rear side (image side) of the projection unit PLU (lens group G1~G5 as imaging projection lens system). Furthermore, the numerical aperture NAo on the object plane (DMD10) side of the projection unit PLU (lens group G1~G5) is represented by the product of the projection magnification Mp and the numerical aperture NAi. When the projection magnification Mp is 1 / 6, it becomes NAo = NAi / 6.

[0037] In the configuration of the illumination unit ILU and projection unit PLU shown in Figures 6 and 4 above, the emitting ends of the fiber bundles FBn (n=1~27) connected to each module MUn (n=1~27) are optically conjugate with the emitting end of the MFE lens 108A of the optical integrator 108 by the input lens system 104. The incident end of the MFE lens 108A is optically conjugate with the center of the mirror surface (central surface) of the DMD10 by the condenser lens system 110. Thus, the illumination light ILm illuminating the entire mirror surface of the DMD10 becomes a uniform illuminance distribution (e.g., intensity non-uniformity within ±1%) by the action of the optical integrator 108. Furthermore, the emitting end of the MFE lens 108A and the pupil Ep of the projection unit PLU are optically conjugate with each other by the condenser lens system 110 and the lens groups G1~G3 of the projection unit PLU.

[0038] Figure 7 is a schematic diagram of the MFE lens 108A of the optical integrator 108 viewed from the exit surface side. The cross-sectional shape of the MFE lens 108A is similar to the overall shape of the mirror surface (image forming area) of the DMD 10, and it is constructed by densely arranging a plurality of lens devices EL with a rectangular cross-section extending in the Y' direction within the X'Y' plane in both the X' and Y' directions. On the incident surface side of the MFE lens 108A, the illumination light ILm from the input lens system 104 shown in Figure 4 forms a roughly circular illumination area Ef. The illumination area Ef has a shape similar to the exit ends of the single or multiple fiber lines of the fiber bundle FB18 (FBn) in Figure 4, and is designed to be a circular area centered on the optical axis AXc.

[0039] In the plurality of lens devices EL of the MFE lens 108A, the point light sources SPFs produced by the illumination light ILm from the emitting end of the fiber bundle FB18 (FBn) are densely distributed in a roughly circular area on the exit surface side of each lens device EL located within the illumination area Ef. Furthermore, the circular area APh in Figure 7 represents the aperture range of a case where an aperture ring with a circular opening is provided on the exit surface side of the MFE lens 108A. The actual illumination light ILm is produced by the plurality of point light sources SPFs dispersed within the circular area APh, and the light from point light sources SPFs outside the circular area APh is blocked.

[0040] Figures 8(A), (B), and (C) are schematic diagrams illustrating an example of the configuration relationship between the point light source SPF formed on the exit surface side of the lens device EL of the MFE lens 108A in Figure 7 and the exit end of the fiber bundle FBn. The coordinate system X'Y' of each of Figures 8(A), (B), and (C) is the same as the coordinate system X'Y' set in Figure 7. Figure 8(A) shows the case where the fiber bundle FBn is a single fiber line, Figure 8(B) shows the case where two fiber lines are arranged in the X' direction as the fiber bundle FBn, and Figure 8(C) shows the case where three fiber lines are arranged in the X' direction as the fiber bundle FBn.

[0041] The emitting end of the fiber bundle FBn and the emitting surface of the MFE lens 108A (lens device EL) are optically set to a conjugate relationship (imaging relationship). Therefore, when the fiber bundle FBn is a single fiber line, as shown in Figure 8(A), a single point light source SPF is formed at the center position on the emitting surface side of the lens device EL. When two fiber lines are bundled together in the X' direction to form the fiber bundle FBn, as shown in Figure 8(B), the geometric centers of the two point light source SPFs are located at the center position on the emitting surface side of the lens device EL. Similarly, when three fiber lines are bundled together in the X' direction to form the fiber bundle FBn, as shown in Figure 8(C), the geometric centers of the three point light source SPFs are located at the center position on the emitting surface side of the lens device EL.

[0042] Furthermore, if the power of the illumination light ILm from the fiber bundle FBn is high, the point light source SPF may cause damage (blurring or burning, etc.) to the lens devices EL of the MFE lens 108A, which is used as an optical integrator, by focusing the light on the surface light source component. In such cases, the focusing position of the point light source SPF can be set in a space slightly offset outward from the exit surface of the MFE lens 108A (the exit surface of the lens device EL). Such a configuration, in which the position of the point light source (focusing point) is offset outward from the lens device by using a compound eye lens illumination system, is disclosed, for example, in U.S. Patent No. 4,939,630.

[0043] Figure 9 schematically illustrates the state of the light source image Ips formed in the pupil Ep within the second lens system 118 of the projection unit PL in Figure 6 when the entire mirror surface of DMD10 is configured as a single plane mirror and this plane mirror is assumed to be tilted at an angle θα / 2 parallel to the tilting mirror 112 in Figure 6. The light source image Ips shown in Figure 9 is formed by re-image of a plurality of point light sources SPF (roughly forming a circular surface light source) formed on the exit surface side of MFE lens 108A. In this case, the single plane mirror configured instead of DMD10 does not produce diffracted or scattered light, and at the center within the pupil Ep, the light source image Ips, which is generated only by positively reflected light (0th order light), is generated only coaxially with the optical axis AXa.

[0044] In Figure 9, when the radius corresponding to the maximum aperture of the pupil Ep is set as re, and the radius corresponding to the effective diameter of the light source image Ips, which is a surface light source, is set as ri, the value σ representing the size (area) of the light source image Ips relative to the size (area) of the pupil Ep is σ = ri / re. The value σ is sometimes appropriately changed to improve the line width or density of the pattern of the projected exposure, or the depth of focus (DOF), etc. The value σ can be changed by setting a variable aperture at the position of the exit surface side of the MFE lens 108A, or at the position of the pupil Ep between the first lens system 116 and the second lens system 118 (in relation to the conjugate of the circular area APh in Figure 7).

[0045] In this type of exposure apparatus EX, the pupil Ep of the projection unit PLU is usually kept at its maximum aperture. Therefore, the change of the σ value is mainly achieved by a variable aperture provided on the exit surface side of the MFE lens 108A. In this case, the radius ri of the light source image Ips is defined by the radius of the circular area APh in Figure 7. Of course, a variable aperture can also be provided on the pupil Ep of the projection unit PLU to adjust the σ value or the depth of focus (DOF).

[0046] However, when the central surface of the DMD10 is set to be perpendicular to the optical axis AXa of the projection unit PLU, and the illumination light ILm is set to a relatively large incident angle θα (e.g., θα ≧ 20°), it has been determined that the intensity distribution of the imaging beam obtained from the reflected light from the micromirror Msa (or Msc) in the on state of the DMD10 at the pupil Ep is not the distribution of the light source image Ips divided by the circular outline as shown in Figure 9, but rather becomes elliptical. This situation will be explained with reference to Figure 10.

[0047] Figure 10 is a simplified representation of the optical path diagram in Figure 6 above, with the orthogonal coordinate system X'Y'Z set to be the same as in Figure 6. Furthermore, for simplicity of explanation, the tilting mirror 112 shown in Figure 6 is omitted. In Figure 10, the tilt angle θd of the micromirror Msa in the open state of DMD10 is set to 17.5° relative to the neutral plane Pcc, which is the design value. Therefore, the angle formed by the optical axis AXb of the MFE lens 108A and the condenser lens system 110 and the optical axis AXa of the projection unit PLU, i.e., the incident angle θα, is set to 35° in the X'Z plane.

[0048] Among the plurality of point light sources SPF formed on the emission side of MFE108A, the illumination lights ILma and ILmb of the two outermost point light sources SPFa and SPFb located in the circular region APh shown in FIG7 are illuminated by the condenser lens system 110 in a plane parallel to the X'Z plane containing the optical axis AXb. The central rays LLa and LLb of the illumination lights ILma and ILmb are parallel to the optical axis AXb until they are incident on the condenser lens system 110. Therefore, when viewed from the DMD10 side, the shape of the surface light source (the collection of point light source SPFs) on the emission side of MFE108A is circular CL1.

[0049] Here, assuming that the reflective surfaces of all the micromirrors of the DMD10 are parallel to the neutral plane Pcc, the illumination lights ILma and ILmb travel along the optical axis AXb', which is tilted at an angle (-θα) symmetrical to the optical axis AXb relative to the optical axis AXa, as normal reflected light. Here, it is assumed that the principal surfaces of the first lens group 116 of the projection unit PLU and the principal surfaces of the condenser lens system 110 are located on the arc Prr centered at the intersection of the neutral plane Pcc of the DMD10 and the optical axis AXa. When viewed from the side of arrow Arw1, the normal reflected light traveling along the optical axis AXb' appears as a circle CL2, the same as the surface light source (the collection of point light sources SPF) on the emission side of the MFE108A.

[0050] However, when viewed from the arrow Arw2 side, which is parallel to the optical axis AXa of the projection unit PLU, the circular surface light source (the collection of point light sources SPF) on the emission side of the MFE108A is estimated to be tilted, so the normally reflected light traveling along the optical axis AXb' appears to be elliptical CL2'. On the other hand, when pattern projection is performed by driving the DMD10, the reflected light (and diffracted light) generated by the multiple open micromirrors Msa becomes the imaging beam Sa' and is incident on the first lens group 116 of the projection unit PLU. Since the first lens group 116 and the condenser lens system 110 are respectively arranged along different optical axes AXa and AXb with tilt angle θα, if one tries to view the intensity distribution of the 0th order light equivalent component (the distribution of the image of the point light source SPF) in the imaging beam Sa' generated by the micromirror Msa in the open state of DMD10 on the pupil Ep, the circular surface light source on the exit surface side of MFE108A will be tilted and estimated, and thus appear to be elliptical CL3.

[0051] When the distribution of the surface light source on the emission surface side of MFE108A is a perfect circle centered on the optical axis AXb, the elliptical CL3 intensity distribution of the imaging beam Sa' (equivalent to the 0th order light component) formed by the pupil Ep of the projection unit PLU is compressed in the incident direction of the illumination light ILm when viewed in the X'Y' plane. The incident direction of the illumination light ILm toward DMD10 is the X' direction in the X'Y' plane, so the major axis of the elliptical CL3 intensity distribution is parallel to the Y' axis, and the minor axis is parallel to the X' axis. If the size of the major axis of the elliptical CL3 intensity distribution is set as Uy', and the size of the minor axis is set as Ux', then the ratio of the ellipse Ux' / Uy' depends on the incident angle θα of the illumination light ILm and becomes cosθα. The incident angle θα is twice the tilt angle θd of the micromirror Msa in the open state of DMD10, so the ratio Ux' / Uy' can also be set by cos(2·θd). When the incident angle θα is set to 35°, the ratio Ux' / Uy' becomes approximately 0.82.

[0052] Figure 11, similar to Figure 9 above, schematically illustrates the situation where the light source image Ips formed in the pupil Ep by the 0th order equivalent component with the highest intensity from the imaging beam Sa from DMD10. For the light source image Ips (elliptical CL3 shape), the radial dimension in the Y' direction becomes the same radius ri as in Figure 9, but the radial dimension in the X' direction becomes a radius ri' that is reduced to about 0.82 times the radius ri. Thus, when the intensity distribution (distribution of the light source image Ips) formed in the pupil Ep by the 0th order equivalent component of the imaging beam Sa is non-isotropic, the imaging characteristics of the edge sometimes differ depending on the direction of the edge of the pattern projected onto the substrate P via the projection unit PLU in the X'Y' plane (i.e., in the XY plane). Therefore, the intensity distribution formed in the pupil Ep by the 0th order equivalent component of the imaging beam Sa is generally ideally set to an isotropic circular shape.

[0053] Therefore, in this embodiment, the circular region APh of the aperture opening shape on the exit surface side of the MFE108A, as described in FIG. 7 above, is transformed into an elliptical region APh' with the X' direction as the major axis and the Y' direction as the minor axis, as shown in FIG. 12. FIG. 12 is a schematic diagram of the MFE lens 108A of the optical integrator 108 viewed from the exit surface side, similar to FIG. 7. The elliptical region APh' is formed by rotating the ellipse CL3 of the light source image Ips formed on the pupil Ep of the projection unit PLU by 90° in the X'Y' plane. Furthermore, the ratio (minor axis dimension / major axis dimension) of the ellipse of the elliptical region APh' is also set to the same cosθα as the ratio of the ellipse CL3 shown in FIG. 10.

[0054] Thus, by setting the effective overall shape (outline) of the surface light source (the collection of point light sources SPF) formed on the exit surface side of the MFE lens 108A to an ellipse, the intensity distribution (light source image Ips) of the zero-order light equivalent component of the imaging beam Sa' formed on the pupil Ep of the projection unit PLU can be set to a circular shape. Regardless of which direction the edge of the pattern extends in the X'Y' plane (in the XY plane), the imaging characteristics (especially the contrast characteristics of the edge) can be made uniform.

[0055] [Telecentricity Error During Projection Exposure] Next, the telecentricity error that may occur when using the exposure apparatus EX of DMD10 as in this embodiment will be explained. Before that, one factor that causes telecentricity error will be briefly explained using FIG13. FIG13(A) and FIG13(B) are schematic diagrams showing the behavior of the imaging beam Sa of the optical path from the pupil Ep shown in FIG6 through the second lens group 118 to the substrate P. The orthogonal coordinate system X'Y'Z in FIG13(A) and FIG13(B) is the same as the coordinate system X'Y'Z in FIG6. For the sake of simplicity, the following scenario is envisioned: the entire mirror surface of DMD10 is set as a plane mirror, tilted at an angle θα / 2 parallel to the tilting mirror 112 in FIG6. In Figures 13(A) and 13(B), lens groups G4 and G5 are arranged along the optical axis AXa between the pupil Ep and the substrate P. An elliptical light source image (surface light source image) Ips is formed inside the pupil Ep, as shown in Figure 11. Furthermore, the principal ray of the reflected light (imaging beam) Sa that passes through a point on the periphery of the light source image (surface light source image) Ips in the X' direction and is incident on the lens groups G4 and G5 is set as La.

[0056] Figure 13(A) shows the behavior of the reflected light (imaging beam) Sa when the center (or centroid) of the light source image (area light source image) Ips is accurately located at the center of the pupil Ep. The principal rays La of the reflected light (imaging beam) Sa directed toward a point within the projection area IAn on the substrate P are all parallel to the optical axis AXa. The imaging beam projected onto the projection area IAn becomes telecentric, that is, the telecentric error is zero. In contrast, Figure 13(B) shows the behavior of the reflected light (imaging beam) Sa when the center (or centroid) of the light source image (area light source image) Ips is shifted laterally by ΔDx in the X' direction from the center of the pupil Ep. In this case, the principal rays La of the reflected light (imaging beam) Sa directed toward a point within the projection area IAn on the substrate P are all tilted by Δθt relative to the optical axis AXa. The tilt amount Δθt becomes the telecentric error. As the tilt amount Δθt (i.e., the lateral displacement ΔDx) exceeds the predetermined allowable value, the imaging state of the pattern image projected onto the projection area IAn decreases.

[0057] [DMD Configuration] As explained above, the DMD10 used in this embodiment is configured for roll and pitch drive. Its specific configuration will be explained with reference to Figures 14 and 15. Figures 14 and 15 are magnified perspective views of a portion of the mirror surface of the DMD10. Here, the orthogonal coordinate system X'Y'Z is the same as the coordinate system X'Y'Z in Figure 6 above. Figure 14 shows the state when the power supply to the drive circuit located beneath each micromirror Ms of the DMD10 is disconnected. In the power-off state, the reflecting surface of each micromirror Ms is set to be parallel to the X'Y' plane. Here, the arrangement spacing of each micromirror Ms in the X' direction is set to Pdx (μm), and the arrangement spacing in the Y' direction is set to Pdy (μm), but in practice, Pdx = Pdy is set.

[0058] Figure 15 shows a situation where the micromirror Msa, in the on state when the power supply to the drive circuit is turned on, and the micromirror Msb, in the off state, coexist. In this embodiment, the micromirror Msa, in the on state, is driven by tilting at an angle θd (=θα / 2) around a line parallel to the Y' axis from the X'Y' plane, and the micromirror Msb, in the off state, is driven by tilting at an angle θd (=θα / 2) around a line parallel to the X' axis from the X'Y' plane. The illumination light ILm illuminates each micromirror Msa and Msb along the principal ray Lp, which is parallel to the X'Z plane (parallel to the optical axis AXb shown in Figure 6). Furthermore, the line Lx' in Figure 15 is formed by projecting the principal ray Lp onto the X'Y' plane and is parallel to the X' axis.

[0059] The incident angle θα of the illumination light ILm toward DMD10 is the tilt angle relative to the Z-axis in the X'Z plane. From the point of view of geometric optics, the micromirror Msa in the open state, tilted at an angle θα / 2 toward the X' direction, produces a reflected light (imaging beam) Sa that travels approximately parallel to the Z-axis in the -Z direction. On the other hand, the reflected light Sg reflected by the micromirror Msb in the closed state is generated in the -Z direction in a state that is not parallel to the Z-axis because the micromirror Msb is tilted toward the Y' direction. In Figure 15, if line Lv is set as a line parallel to the Z-axis (optical axis AXa), and line Lh is set as the projection of the principal ray of the reflected light Sg onto the X'Y' plane, then the reflected light Sg travels in the tilt direction in the plane containing lines Lv and Lh.

[0060] [Imaging State Obtained by DMD] In projection exposure using DMD10, while rapidly switching the tilt of the multiple micromirrors Ms between their on and off states based on pattern data (drawing data) as shown in FIG15, the substrate P is scanned and moved in the X direction at a speed corresponding to the switching speed to perform pattern exposure. However, depending on the fineness or density, or periodicity of the projected pattern, the telecentricity of the imaging beam projected onto the substrate P from the projection unit PLU (first lens group 116 and second lens group 118) sometimes varies. This is because, depending on the different tilt states of the multiple micromirrors Ms of DMD10 corresponding to the pattern, the mirror surface of DMD10 sometimes functions as a reflective diffraction grating (blazed diffraction grating).

[0061] Figure 16 shows a portion of the mirror surface of DMD10 viewed in the X'Y' plane, and Figure 17 shows the a-a' arrow view of the mirror surface of DMD10 in Figure 16 viewed in the X'Z plane. In Figure 16, only one row of micromirrors Ms arranged along the Y' direction is in the open state (Msa), while the other micromirrors Ms are in the closed state (Msb). The tilted state of the micromirrors Ms in Figure 16 occurs when there is an isolated line pattern with a linewidth of approximately 1 μm at the projection resolution limit. In the X'Y' plane, the reflected light (imaging beam) Sa from the open state micromirror Msa is generated parallel to the Z-axis in the -Z direction, while the reflected light Sg from the closed state micromirror Msb is in the -Z direction, but tilted along the line Lh in Figure 16.

[0062] In this case, as shown in Figure 17, only one of the plurality of micromirrors Ms arranged in the X' direction becomes an open micromirror Msa with an angle θd (=θα / 2) tilted relative to the neutral plane Pcc (a plane parallel to the X'Y' plane containing the center point of all micromirrors Ms) about a line parallel to the Y' axis. Therefore, if viewed in the X'Z plane, the reflected light (imaging beam) Sa generated by the open micromirror Msa becomes a simple normal reflected light without diffraction of more than one order, and its principal ray La is parallel to the optical axis AXa and incident on the projection unit PLU. The reflected light Sg from the other closed micromirrors Msb does not incident on the projection unit PLU. Furthermore, when the open micromirror Msa is an isolated one in the X' direction (or a column arranged in the Y' direction), the principal ray La of the reflected light (imaging beam) Sa is independent of the wavelength λ of the illumination light ILm and is parallel to the optical axis AXa.

[0063] Figure 18 is a schematic diagram in the X'Z plane showing the imaging state of the reflected light (imaging beam) Sa from the isolated micromirror Msa as shown in Figure 17, obtained by the projection unit PLU. In Figure 18, components with the same function as those described in Figure 6 above are labeled with the same symbols. The projection unit PLU (lens group G1 to G5) is a telecentric reduction projection system. Therefore, if the principal ray La of the reflected light (imaging beam) Sa from the isolated micromirror Msa is parallel to the optical axis AXa, then the principal ray La of the reflected light (imaging beam) Sa that forms the reduced image ia is also parallel to the perpendicular line (optical axis AXa) of the surface of the substrate P, and no telecentric error occurs. Furthermore, the numerical aperture NAo of the reflected light (imaging beam) Sa on the object side (DMD10) side of the projection unit PLU shown in Figure 18 is equal to the numerical aperture of the illumination light ILm.

[0064] As explained in Figures 11 (or 9) and 13(A) above, when the DMD10 is set as a large plane mirror with an angle of θα / 2, the center (centroid) of the light source image (surface light source image) Ips formed in the pupil Ep of the projection unit PLU passes through the optical axis AXa. Similarly, when the normal reflection light Sa from the isolated micromirror Msa in the mirror of the DMD10 is incident on the projection unit PLU, since the reflecting surface of the micromirror Ms is a fine rectangle (square), the point image intensity distribution of the normal reflection light Sa at the position (Fourier transform surface) of the beam Isa at the pupil Ep is represented by a sinc2 function (point image intensity distribution of a square opening) centered on the optical axis AXa.

[0065] Figure 19 is a schematic diagram showing the theoretical point image intensity distribution Iea (made from the beam of light from a single point source SPF shown in Figures 7 and 8) of the beam of light Sa reflected from a single micromirror Msa isolated in the X' direction (here, the 0th order diffracted light) at the pupil Ep. In the diagram of Figure 19, the horizontal axis represents the coordinate position in the X' (or Y') direction, which is set as the optical axis AXa, and the vertical axis represents the light intensity Ie. The point image intensity distribution Iea is represented by the following equation (1).

[0066] [Number 1] Ie=Io·sinc2(X')=Io·sin2(X') / (X')2···(1)

[0067] In this formula (1), Io represents the peak value of the light intensity Ie. The position of the peak value Io obtained from the reflected light Sa from an isolated column (or single) micromirror Msa is consistent with the position of the origin O in the X' (or Y') direction, i.e., the position of the optical axis AXa. Furthermore, as explained in Figure 12 above, when the shape of the surface light source formed on the emission surface side of MFE108A is adjusted to be like an elliptical region APh, the position ±ra of the first dark line in the X' (or Y') direction where the light intensity Ie of the point image intensity distribution Iea initially reaches its minimum value (0) from the origin O roughly corresponds to the position of the radius ri of the light source image Ips as explained in Figure 9 above. Moreover, the actual intensity distribution at the pupil Ep is obtained by integrating (convolution operation) the point image intensity distribution Iea over the extended range (σ value) of the light source image Ips shown in Figure 9, resulting in a roughly uniform intensity.

[0068] Next, referring to Figures 20 and 21, the case where the width of the projected pattern in the X' direction (X direction) is sufficiently large will be explained. Figure 20 is a diagram showing a portion of the mirror surface of DMD10 viewed in the X'Y' plane, and Figure 21 is a diagram showing the a-a' arrow view of the mirror surface of DMD10 in Figure 20 viewed in the X'Z plane. Figure 20 shows the case where all of the plurality of micromirrors Ms shown in Figure 16 above are in the open state. In Figure 20, only the arrangement of 9 micromirrors Ms in the X' direction and 10 micromirrors Ms in the Y' direction is shown, but sometimes a larger number of adjacent micromirrors Ms (or all the micromirrors Ms on DMD10) are in the open state.

[0069] As shown in Figures 20 and 21, a plurality of micromirrors Msa arranged in an open state adjacent to each other in the X' direction generate reflected light Sa' by diffraction in a state slightly tilted from the optical axis AXa. If the mirror surface of DMD10 in the state of Figure 21 is considered as a diffraction grating arranged along the neutral plane Pcc in the X' direction with a spacing Pdx, then j is set as the order (j=0, 1, 2, 3, ···), λ is set as the wavelength, and the incident angle of the illumination light ILm is set as θα. The generation angle θj of the diffracted light is expressed as shown in the following formula (2) or formula (3).

[0070] [Number 2] sinθj=j(λ / Pdx)-sinθα···(2)

[0071] [Number 3] sinθj=sinθα-j(λ / Pdx)···(3)

[0072] Figure 22 is an example showing the distribution of the angle θj of the diffracted light Idj calculated when the incident angle θα of the illumination light ILm (the tilt angle of the principal ray Lp of the illumination light ILm relative to the optical axis AXa) is set to 35.0°, the tilt angle θd of the micromirror Msa in the open state is set to 17.5°, the distance Pdx between the micromirrors Msa is set to 5.4 μm, and the wavelength λ is set to 355.0 nm. As shown in Figure 22, the incident angle θα of the illumination light ILm is 35°, so the 0th order diffracted light Id0 (j=0) is tilted to +35° relative to the optical axis AXa. As the diffraction order increases, the angle θj relative to the 0th order diffracted light Id0 increases. The values ​​shown in the lower part of Figure 22 represent the order j in parentheses and the tilt angle of the diffracted light Idj of each order from the optical axis AXa.

[0073] Under the numerical conditions shown in Figure 22, the tilt angle of the 9th diffracted beam Id9 from the optical axis AXa is the smallest, becoming approximately -1.04°. Therefore, when the micromirrors Ms of DMD10 are densely in the open state as shown in Figures 20 and 21, the center of the intensity distribution of the imaging beam (Sa') in the pupil EP of the projection unit PLU is offset from the position of the optical axis AXa by an angle equivalent to -1.04° (equivalent to the lateral displacement ΔDx shown in Figure 13(B) above). The actual distribution of the imaging beam in the pupil Ep is obtained by performing a convolution operation on the diffracted beam distribution represented by Equation (2) or (3) using the sinc2 function represented by Equation (1).

[0074] Figure 23 is a schematic diagram showing the intensity distribution of the imaging beam Sa' at the pupil Ep when the diffracted light is generated as shown in Figure 22. The horizontal axis in Figure 23 represents the value obtained by converting the angle θj of the diffracted light Idj into the numerical aperture NAo on the object plane (DMD10) side and the numerical aperture NAi on the image plane (substrate P) side when the projection magnification Mp of the projection unit PLU is set to 1 / 6. It is also assumed that the numerical aperture NAi on the image plane side of the projection unit PLU is 0.3 (the numerical aperture NAo on the object plane side is 0.05). In this case, the resolution (minimum resolving linewidth) Rs is expressed by Rs = k1(λ / NAi) using the process constant k1 (0 < k1 ≦ 1).

[0075] Therefore, the resolving power Rs at wavelength λ = 355.0 nm and k1 = 0.7 becomes approximately 0.83 μm. The distance Pdx (Pdy) between micromirrors Ms is reduced to 0.9 μm on the image plane (substrate P) side by a projection magnification Mp = 1 / 6. Therefore, if the projection unit Plu has an image plane numerical aperture NAI of 0.3 or higher (object plane numerical aperture NAO of 0.05 or higher), a single projected image of a micromirror Msa in the open state can be imaged with high contrast.

[0076] In Figure 23, the maximum aperture of the pupil Ep of the projection unit PLU, i.e., the numerical aperture NAo = 0.05 on the object side, is θe at the angle θe from the optical axis AXa in the X' direction, which is approximately ±2.87° according to NAo = sinθe. As shown in Figure 22 above, the tilt angle of the 9th diffracted light Id9, -1.04° (more precisely -1.037°), when converted to the numerical aperture NAo on the object side, becomes approximately 0.018. The intensity distribution Hpa of the imaging beam Sa' (equivalent to the 0th order light component) at the pupil Ep is shifted in the X' direction by a displacement ΔDx from the original position of the image Ips (radius ri) of the light source. Furthermore, a portion of the intensity distribution Hpb obtained by the 8th diffracted light Id8 also appears in the periphery of the +X' direction within the pupil Ep, but its peak intensity is low. Furthermore, the tilt angle of the 10th diffracted light Id10 on the object side from the optical axis AXa is as large as 4.81°, so its intensity distribution is distributed outside the pupil Ep and does not pass through the projection unit PLU. Moreover, the intensity distributions Hpa and Hpb in Figure 23, as explained in Figure 12 above, become approximately circular by setting the surface light source formed on the emission surface side of the MFE108A of the illumination unit ILU as an elliptical region APh'.

[0077] Furthermore, the micromirrors Ms of the DMD10 are also arranged at a spacing Pdy (=5.4 μm) in the Y' direction. Therefore, corresponding to this spacing Pdy, they also generate diffracted light with low illumination in the Y' direction, producing weak intensity distributions Hpc and Hpd. The intensity distributions Hpc and Hpd vary depending on the size of the numerical aperture NAo (NAi) of the projection unit PLU, and sometimes a portion of them enters the pupil Ep. Therefore, by appropriately setting the relationship between the numerical aperture NAo (NAi) of the projection unit PLU and the size (radius ri) of the light source image Ips, the intensity distributions Hpc and Hpd can be prevented from entering the pupil Ep.

[0078] As illustrated in Figure 13(B) above, the telecentric error Δθt on the image plane side caused by the shift ΔDx of the center of the intensity distribution Hpa becomes Δθt = -6.22° (= -1.037° / projection magnification Mp) under the conditions shown in Figures 22 and 23. Thus, when the majority of the micromirrors Ms in the DMD10 are exposed to form a large pattern that is in an open state, the principal ray of the imaging beam Sa' towards the substrate P is tilted by more than 6° relative to the optical axis AXa. This telecentric error Δθt also contributes to the sometimes reduced imaging quality (contrast characteristics, distortion characteristics, symmetry, etc.) of the projected image.

[0079] Next, referring to Figures 24 and 25, the case where the projected pattern is a pattern of lines and gaps with a certain spacing in the X' direction (X direction) will be explained. Figure 24 is a diagram showing a portion of the mirror surface of DMD10 viewed in the X'Y' plane, and Figure 25 is a diagram showing the a-a' arrow view of the mirror surface of DMD10 in Figure 24 viewed in the X'Z plane. Figure 24 shows the case where, among the plurality of micromirrors Ms shown in Figure 16 above, the odd-numbered micromirrors Msa arranged in the X' direction are in the open state, and the even-numbered micromirrors Msb are in the closed state. Assume that the odd-numbered micromirrors Ms in the X' direction are all in the open state in a row arranged in the Y' direction, and the even-numbered micromirrors Ms are all in the closed state in a row arranged in the Y' direction.

[0080] As shown in Figure 25, when the micromirrors Msa are arranged one after another along the X' direction in the open state, the mirror surface of DMD10 is considered as a diffraction grating arranged along the neutral plane Pcc in the X' direction with a spacing of 2·Pdx. The generation angle θj of the diffracted light generated from DMD10 is represented by the following formula (4) or formula (5), which is the same as the above formula (2) or formula (3).

[0081] [Number 4] sinθj=j(λ / 2Pdx)-sinθα···(4)

[0082] [Number 5] sinθj=sinθα-j(λ / 2Pdx)···(5)

[0083] Figure 26, similar to Figure 22, shows the distribution of the angle θj of the diffracted light Idj calculated with the incident angle θα of the illumination light ILm (the tilt angle of the principal ray Lp of the illumination light ILm relative to the optical axis AXa) set to 35.0°, the tilt angle θd of the micromirror Msa in the open state set to 17.5°, the distance 2Pdx between the micromirrors Msa set to 10.8 μm, and the wavelength λ set to 355.0 nm. As shown in Figure 26, the incident angle θα of the illumination light ILm is 35°, so the 0th order diffracted light Id0 (j=0) is tilted to +35° relative to the optical axis AXa. As the diffraction order increases, the angle θj relative to the 0th order diffracted light Id0 increases. The values ​​shown in the lower part of Figure 26 represent the order j in parentheses and the tilt angle of the diffracted light Idj of each order from the optical axis AXa.

[0084] Under the numerical conditions shown in Figure 26, the tilt angle of the 17th diffracted beam Id17 from the optical axis AXa is the smallest, approximately 0.85°. Furthermore, the 18th diffracted beam Id18 with a tilt angle of -1.04° from the optical axis AXa is also generated. Therefore, when the micromirror Ms of the DMD10 is in the open state with the finest line and gap as shown in Figures 24 and 25, the center of the intensity distribution of the imaging beam Sa' in the pupil EP of the projection unit PLU is offset from the position of the optical axis AXa by an angle equivalent to 0.85° or -1.04°. The actual distribution of the imaging beam Sa' in the pupil EP is obtained by performing a convolution operation on the diffracted beam distribution represented by equation (4) or equation (5) using the sinc2 function represented by equation (1).

[0085] The situation in Figure 26 is similar to that in Figure 23 above. The intensity distribution Hpa of the imaging beam (normal reflection component) at the pupil Ep corresponds to the tilt angle of 0.85° for the 17th diffraction Id17 and the tilt angle of -1.04° for the 18th diffraction Id18, respectively, which is caused by the original position of the image Ips (radius ri) of the light source shifting in the X' direction. In the case of the diffraction distribution as shown in Figure 26, the intensity distribution Hpa formed in the direction of the 17th diffraction Id17 and the intensity distribution Hpa formed in the direction of the 18th diffraction Id18 are of different strengths. Therefore, the telecentric error Δθt on the image plane side caused by the shift of the intensity distribution Hpa is approximately within the range of Δθt = 5.1° and Δθt = -6.22°.

[0086] This range is slightly different from the case where multiple micromirrors Ms are adjacent to each other and micromirrors Msa are in the open state as shown in Figures 20 and 21 above, where the telecentric error Δθt = -6.22° is the generation direction of the 9th diffraction light Id9 (see Figure 22). Furthermore, it is quite different from the case where one column (or a single micromirror) of multiple micromirrors Ms is isolated and in the open state as shown in Figures 16 and 17 above, where the telecentric error Δθt = 0° is the same. Moreover, the actual pattern image projected onto the substrate P by the projection unit PLU is formed by the interference of the reflected light Sa' containing the diffracted light from DMD10, which is introduced into the projection unit PLU. Furthermore, Equation (4) or Equation (5) can be used to determine the generation state of diffracted light in a pattern of lines and gaps with an arrangement spacing or line width of n times Pdx (5.4 μm) by setting n to a real number in Equation (6) or Equation (7).

[0087] [Number 6] sinθj=j(λ / (n·Pdx))-sinθα···(6)

[0088] [Number 7] sinθj=sinθα-j (λ / (n·Pdx))···(7)

[0089] Figure 27 is a schematic diagram corresponding to Figure 23 above, showing the distribution of the reflected light (diffracted light) from DMD10 at the pupil Ep of the projection unit PLU as shown in Figure 26. As explained in Figure 12 above, the situation in Figure 27 is such that by setting the outline of the surface light source formed on the emission surface side of MFE108A to an elliptical shape APh', the intensity distribution of the imaging beam Sa' formed on the pupil Ep of the projection unit PLU and the diffracted beam becomes circular. Furthermore, in Figure 27, assuming that the intensity of the 18th order diffracted light Id18 shown in Figure 26 is the largest, in the case of the projection of the line and gap pattern as shown in Figures 24 and 25, the intensity distribution of the 18th order diffracted light Id18 is taken as the intensity distribution Hpa, which is equivalent to the 0th order component. The intensity distribution Hpa corresponds to the angle of -1.04° from the optical axis AXa of the 18th diffracted light Id18, and is eccentrically oriented towards the -X' direction by ΔDx, resulting in a telecentric error Δθt.

[0090] As illustrated in Figure 23 above, within the plane of the pupil Ep, intensity distributions Hpb, Hpc, and Hpd of the diffracted light components are generated due to the arrangement spacing Pdx and Pdy of the micromirrors Ms in the X' and Y' directions of the DMD10, but their intensity is sufficiently small compared to the intensity distribution Hpa. Furthermore, based on the line and gap pattern (the line width in the X' direction is Pdx and the spacing is 2Pdx) produced by the micromirrors Ms of the DMD10, the intensity distribution ±Hpb' of the ±1st order light equivalent components generated by diffraction appears on both sides of the intensity distribution Hpa in the X' direction. The center point PXp of the intensity distribution +Hpb' of the +1st order light equivalent component is located approximately midway between the center point (Id18) of the intensity distribution Hpa of the 0th order light equivalent component and the center point of the intensity distribution Hpb in the +X' direction. Similarly, the center point PXm of the intensity distribution -Hpb' of the -1 light equivalent component is located approximately midway between the center point (Id18) of the intensity distribution Hpa of the 0 light equivalent component and the center point of the intensity distribution Hpb in the -X' direction.

[0091] Furthermore, Figure 27 shows the intensity distribution of the imaging beam Sa' (diffraction beam) at the pupil Ep, which is a pattern of lines and gaps with a spacing of 2Pdx in the X' direction, as shown in Figure 24. In contrast, when the pattern is a pattern of lines and gaps with a spacing of 2Pdy (Pdy = Pdx) in the Y' direction, the intensity distribution ±Hpb' of the ±1st order light equivalent component appears on both sides of the intensity distribution Hpa in the Y' direction, with the center point (Id18) of the intensity distribution Hpa of the 0th order light equivalent component eccentricated by ΔDx in the -X' direction.

[0092] Thus, when most of the micromirrors Ms in the DMD10 are in an open state with lines and gaps, the principal ray of the imaging beam toward the substrate P is sometimes significantly tilted relative to the optical axis AXa, which sometimes significantly reduces the imaging quality (contrast characteristics, distortion characteristics, etc.) of the projected image. Therefore, an example of the change in imaging quality caused by the telecentric error Δθt will be explained with reference to FIG28. FIG28 is a graph showing the result of simulating a spatial image of a line and gap pattern with a line width of 1 μm and a distance of 2 μm in the X' direction on the image plane. The horizontal axis of FIG28 represents the position (μm) in the X' direction on the image plane, and the vertical axis represents the relative intensity value obtained by normalizing the intensity of the illumination light (incident light) to 1.

[0093] In the graph of Figure 28, the numerical aperture NAi of the image side of the projection unit PLU is set to 0.25, the σ value of the illumination light ILm is set to 0.6, and the imaging beam Sa' at the pupil Ep of the projection unit PLU is eccentric in the X' direction relative to the optical axis AXa, with the telecentric error Δθt on the image side being 50 mrad (≈2.865°) for simulation. In the graph of Figure 28, the characteristic Q1 shown by the dashed line is the contrast characteristic of the optimal focal plane (optimal imaging plane) of the projection unit PLU, and the characteristic Q2 shown by the solid line is the contrast characteristic of the plane defocused by 3 μm from the optimal focal plane in the direction of the optical axis AXa. Furthermore, in Figure 28, dark lines with a line width of 1 μm are formed at positions 0, ±2 μm, and ±4 μm, for a total of 5 locations.

[0094] Typically, the contrast (intensity amplitude) of characteristic Q2 is lower than that of characteristic Q1 due to defocusing. However, it is known that the symmetry between the characteristics near +5 μm and the characteristics near -5 μm is degraded due to the influence of the telecentric error Δθt. According to this situation, when the telecentric error Δθt on the image plane side exceeds the allowable range (e.g., ±2°), that is, when the micromirrors Msa in the open state of the multiple micromirrors Ms of DMD10 are densely or periodically arranged over a wide range, the accuracy of the edge position of the resist image corresponding to the edge portion of the exposed pattern is impaired, resulting in errors in the linewidth or size of the pattern. That is, the intensity distribution (distribution of diffracted light) of the pupil Ep of the projection unit PLU formed by the reflected light (imaging beam) Sa' from DMD10 deviates from the isotropic state or symmetrical state centered on the optical axis AXa. Accompanying this, the asymmetry of the projected pattern image increases.

[0095] [Wavelength Dependence of Telecentric Error] The telecentric error Δθt described above, as shown by equations (2) to (5) above, varies depending on the wavelength λ. For example, in the case of the states in Figures 20 and 21 represented by equation (2), in order to set the telecentric error Δθt on the image plane side to zero, it is only necessary to set the wavelength λ such that the tilt angle of the 9th diffracted light Id9 from the optical axis AXa is -1.04° (more precisely -1.037°) to zero.

[0096] Figure 29 is a graph showing the relationship between the center wavelength λ and the telecentric error Δθt based on equation (2) above. The horizontal axis represents the center wavelength λ (nm), and the vertical axis represents the telecentric error Δθt (deg) on ​​the image plane side. When the distance between the micromirrors Ms of the DMD10 is set to 5.4 μm, the tilt angle θd of the micromirrors Ms is set to 17.5°, and the incident angle θα of the illumination light ILm is set to 35°, and the micromirrors Ms are densely arranged in the open state as shown in Figures 20 and 21, the telecentric error Δθt theoretically becomes zero when the center wavelength λ is about 344.146 nm. Ideally, the telecentric error Δθt on the image plane side should be set to zero, but it can be within an acceptable range depending on the minimum linewidth (or resolution Rs) of the pattern to be projected.

[0097] For example, when the permissible range of the telecentric error Δθt on the image plane side is set to within ±0.6° (approximately 10 mrad), as shown in Figure 29, the center wavelength λ only needs to be in the range of 343.098 nm to 345.193 nm (with a width of 2.095 nm). Furthermore, when the permissible range of the telecentric error Δθt on the image plane side is set to within ±2.0°, the center wavelength λ only needs to be in the range of 340.655 nm to 347.636 nm (with a width of 6.98 nm).

[0098] Thus, the telecentric error Δθt caused by the arrangement (periodicity) or density, i.e., the distribution density, of the micromirrors Msa in the DMD10 when it is in the on state is also wavelength-dependent. Generally, the specifications such as the distance between the micromirrors Ms Pdx (Pdy) or the tilt angle θd of the DMD10 are uniquely set as finished products (e.g., ultraviolet-compatible DMDs manufactured by Texas Instruments), and therefore the wavelength λ of the illumination light ILm is set in a manner that conforms to its specifications. In this embodiment, the distance between the micromirrors Ms Pdx (Pdy) of the DMD10 is set to 5.4 μm and the tilt angle θd is set to 17.5°, so as a light source that supplies illumination light ILm to each fiber bundle FBn (n=1~27), ​​a fiber amplifier laser light source that generates high-brightness ultraviolet pulse light can be used.

[0099] The fiber optic amplifier laser source, for example, as disclosed in Patent No. 6428675, comprises the following components: a semiconductor laser device that generates seed light in the infrared wavelength domain; a high-speed switching device (electro-optical device, etc.) for the seed light; an optical fiber that amplifies the switched seed light (pulse light) using pump light; and a wavelength conversion device that converts the amplified infrared wavelength domain light into high harmonic (ultraviolet wavelength domain) pulse light. In the case of this fiber optic amplifier laser source, the peak wavelength of ultraviolet light with a generation efficiency (conversion efficiency) of 343.333 nm can be improved by combining the available semiconductor laser device, optical fiber, and wavelength conversion device. At this peak wavelength, the maximum image-plane side telecentric error Δθt (the tilt angle of the 9th diffracted light Id9 on the image-plane side in Figures 22 and 23) that may occur in the state of FIG20 becomes approximately 0.466° (approximately 8.13 mrad).

[0100] Based on the above, when combining two or more lights with significantly different peak wavelengths (e.g., light with a wavelength of more than 350 nm and light with a wavelength of more than 400 nm) as illumination light ILm, the telecentric error Δθt may vary considerably depending on the shape of the pattern to be projected (isolated pattern, line and gap pattern, or large land-like pattern). In this embodiment, the illumination light ILm supplied to each module MUn (n = 1 to 27) is obtained by combining and using light from a plurality of fiber amplifier laser sources whose peak wavelengths are slightly offset within the allowable range of wavelength-dependent telecentric error Δθt. By using the illumination light ILm obtained by combining a plurality of lights with slightly offset peak wavelengths in this way, the contrast of spots (or interference fringes) generated on the micromirror Ms of DMD10 (and on the substrate P) can be suppressed by the interferometry of the illumination light ILm.

[0101] [Telecentric Adjustment Mechanism] As explained above, when the micromirrors Ms in the plurality of micromirrors of the DMD10 are arranged closely in the X' and Y' directions according to the pattern to be exposed on the substrate P, or arranged periodically in the X' (or Y') direction, the imaging beam (Sa, Sa') projected from the projection unit PLU will produce a telecentric error (angle change) Δθt of varying magnitude. Each of the plurality of micromirrors Ms in the DMD10 switches between the on and off states with a response speed of about 10 kHz, so the pattern image generated by the DMD10 also changes rapidly according to the depicted data. Therefore, during the scanning exposure of the pattern on the display panel, etc., the pattern image projected from each module MUn (n=1~27) instantaneously changes shape into isolated line or dot patterns, line and gap patterns, or large shore-like patterns, etc.

[0102] A typical television display panel (liquid crystal type, organic EL (electroluminescence) type) consists of an image display area formed by arranging pixels of approximately 200-300 μm square on a substrate P in a matrix with a predetermined aspect ratio such as 2:1 or 16:9, and peripheral circuitry (extended wiring, connecting pads, etc.) disposed around it. Within each pixel, a thin-film transistor (TFT) for switching or current driving is formed. However, the size (linewidth) of the TFT pattern (patterns of gate layer, drain / source layer, semiconductor layer, etc.) or gate wiring or driving wiring is sufficiently small compared to the pixel spacing (200 μm-300 μm). Therefore, when the pattern within the image display area is exposed, the pattern image projected from the DMD10 is essentially isolated, thus eliminating telecentricity error Δθt.

[0103] However, depending on the configuration of the lamp driving circuit (TFT circuit) of each pixel, sometimes lines and gaps are formed in the X or Y direction with a spacing smaller than the pixel spacing. In this case, when the pattern in the image display area is exposed, the pattern image projected from the DMD10 is periodic. Therefore, depending on the degree of periodicity, a telecentric error Δθt may occur. Furthermore, during the exposure of the image display area, sometimes a rectangular pattern that is approximately the same size as the pixel or more than half the size of the pixel is uniformly exposed. In this case, more than half of the micromirrors Ms of the DMD10 in the exposure of the image display area are in a basically dense state when on. Therefore, a relatively large telecentric error Δθt may occur.

[0104] The state of the telecentric error Δθt can be estimated before exposure based on the drawing data of the pattern used in the display panel exposed by each of the plurality of modules MUn (n = 1 to 27). In this embodiment, the position or orientation of each of the optical components in the module MUn is configured to be micro-adjustable. Among these optical components, an adjustable optical component can be selected to correct the telecentric error Δθt according to the estimated magnitude of the telecentric error Δθt.

[0105] Figure 30 shows the specific configuration of the optical path from the fiber bundle FBn in the illumination unit ILU of the module MUn shown in Figure 4 or Figure 6 above to the MFE lens 108A. Figure 31 shows the specific configuration of the optical path from the MFE lens 108A in the illumination unit ILU to the DMD10. In Figures 30 and 31, the orthogonal coordinate system X'Y'Z is set to be the same as the coordinate system X'Y'Z in Figure 4 (Figure 6), and the same symbols are used to label components with the same function as those shown in Figure 4.

[0106] Figure 4 is omitted, but in Figure 30, a contact lens 101 is arranged immediately after the emitting end of the fiber bundle FBn to suppress the propagation of the illumination light ILm from the emitting end. The optical axis of the contact lens 101 is set to be parallel to the Z-axis. The illumination light ILm traveling from the fiber bundle FBn with a predetermined numerical aperture is reflected by the mirror 100, travels parallel to the X' axis, and is reflected by the mirror 102 in the -Z direction. The input lens system 104 arranged in the optical path from the mirror 102 to the MFE lens 108A consists of three lens groups 104A, 104B, and 104C that are spaced apart from each other along the optical axis AXc.

[0107] The illuminance adjustment filter 106 is supported on a holding member 106A that is moved by a drive mechanism 106B and is disposed between the lens group 104A and the lens group 104B. One example of the illuminance adjustment filter 106 is as disclosed in, for example, Japanese Patent Application Publication No. 11-195587, in which a fine light-blocking dot pattern or a multi-row of elongated light-blocking wedge-shaped patterns is formed by gradually changing the density on a quartz or other light-transmitting plate. By moving the quartz plate in parallel, the transmittance of the illumination light ILm can be continuously varied within a predetermined range.

[0108] The first telecentric adjustment mechanism is composed of the following parts: a tilting mechanism 100A, which makes a two-dimensional tilt (rotation angle about the X' axis and about the Y' axis) of the mirror 100 that reflects the illumination light ILm from the optical fiber bundle FBn; a translation mechanism 100B, which makes the mirror 100 move two-dimensionally in the X'Y' plane perpendicular to the optical axis AXc; and a drive unit 100C using a micro-head or piezoelectric actuator, which drives the tilting mechanism 100A and the translation mechanism 100B respectively.

[0109] By adjusting the tilt of the adjustment mirror 100, the central ray (principal ray) of the illumination light ILm incident on the input lens system 104 can be adjusted to be coaxial with the optical axis AXc. Furthermore, the exit end of the fiber bundle FBn is positioned at the front focal point of the input lens system 104. Therefore, if the mirror 100 is moved slightly in the X' direction, the central ray (principal ray) of the illumination light ILm incident on the input lens system 104 moves parallel to the optical axis AXc in the X' direction. Thus, the central ray (principal ray) of the illumination light ILm emitted from the input lens system 104 travels at a slight tilt relative to the optical axis AXc. Therefore, the illumination light ILm incident on the MFE lens 108A is slightly tilted overall in the X'Z plane.

[0110] Figure 32 is an exaggerated representation of the state of the point light source SPF formed on the exit surface side of the MFE lens 108A when the illumination light ILm incident on the MFE lens 108A is tilted in the X'Z plane. When the central ray (principal ray) of the illumination light ILm is parallel to the optical axis AXc, the point light source SPF that focuses the light on the exit surface side of each lens device EL of the MFE lens 108A is shown as the white dot in Figure 32, located at the center in the X' direction. If the illumination light ILm is tilted relative to the optical axis AXc in the X'Z plane, the point light source SPF that focuses the light on the exit surface side of each lens device EL is shown as the black dot in Figure 32, eccentric from the center position in the X' direction by Δxs. In this case, as illustrated in Figures 7 to 9 above, the surface light source obtained by the assembly of a plurality of point light sources SPF formed on the exit surface side of the MFE lens 108A is transversely shifted by Δxs in the X' direction. The cross-sectional dimensions of each lens element EL of the MFE lens 108A in the X'Y' plane are small, so the eccentricity Δxs of the surface light source in the X' direction is also small.

[0111] As shown in FIG30, an aperture ring 108B with an opening shape of an elliptical region APh' as ​​shown in FIG12 is provided on the exit surface side of the MFE lens 108A. The MFE lens 108A and the aperture ring 108B are integrally mounted on the holding part 108C. The holding part 108C (MFE 108A) is provided in such a way that its position in the X'Y' plane can be finely adjusted by means of a micro-motion mechanism 108D using a micro-head or a piezoelectric motor. In this embodiment, the micro-motion mechanism 108D that makes the MFE lens 108A move two-dimensionally in the X'Y' plane functions as a second telecentric adjustment mechanism. As shown in FIG31, the aperture ring 108B has an opening of an elliptical region APh' with the X' direction as the major axis and the Y' direction as the minor axis. If the major axis of the elliptical region APh' is set as Ux and the minor axis as Uy, then the ratio Uy / Ux of the ellipse depends on the cosine of the incident angle θα of the illumination light ILm toward DMD10 (twice the tilt angle θd of the micromirror Msa in the open state), and is set as Uy / Ux = cosθα.

[0112] Immediately following the MFE lens 108A (aperture 108B), a plate-shaped beam splitter 109A is provided, tilted at approximately 45° relative to the optical axis AXc. The beam splitter 109A allows most of the illumination light ILm from the MFE lens 108A to pass through, while the remaining light (e.g., a few percent) is reflected towards the condenser lens 109B. A portion of the illumination light ILm, focused by the condenser lens 109B, is guided by the fiber bundle 109C to the photoelectric device 109D. The photoelectric device 109D is used as an integral sensor (cumulative monitoring) to monitor the intensity of the illumination light ILm and measure the exposure of the imaging beam projected onto the substrate P.

[0113] As shown in Figure 31, the illumination light ILm from the surface light source (an assembly of point light sources SPF) on the exit surface side of the MFE lens 108A passes through the beam splitter 109A and enters the condenser lens system 110. The condenser lens system 110 is composed of a front group lens system 110A and a rear group lens system 110B arranged with open spacing. Its two-dimensional position in the X'Y' plane can be finely adjusted by a micro-motion mechanism 110C using a micro-head or a piezoelectric motor. That is, the eccentricity adjustment of the condenser lens system 110 can be performed by the micro-motion mechanism 110C. In this embodiment, the micro-motion mechanism 110C, which makes the condenser lens system 110 move two-dimensionally in the X'Y' plane, functions as a third telecentric adjustment mechanism. Furthermore, the first, second, and third telecentric adjustment mechanisms are all adjusted to the relative positional relationship between the surface light source generated on the exit surface side of the MFE lens 108A (or the surface light source confined within the opening of the elliptical region APh' of the aperture 108B) and the condenser lens system 110, which is related to the direction of eccentricity.

[0114] The front focal point of the condenser lens system 110 is set at the position of the surface light source (the assembly of point light source SPFs) on the exit surface side of the MFE lens 108A. The illumination light ILm, which travels telecentrically from the condenser lens system 110 via the tilting mirror 112, illuminates the DMD 10 with Köhler light. As explained in Figure 32 above, if the surface light source as a whole, which is the assembly of a plurality of point light source SPFs formed on the exit surface side of the MFE lens 108A, is laterally shifted by Δxs in the X' direction, the principal ray (central ray) of the illumination light ILm illuminating the DMD 10 becomes slightly tilted relative to the optical axis AXb in Figure 31. That is, by using the first telecentric adjustment mechanism to deliberately impart a telecentric error to the illumination light ILm, the incident angle θα of the illumination light ILm illustrated in Figures 6, 17, 21 and 25 above can be slightly changed in the X'Z plane from the initial set angle (35.0°).

[0115] Furthermore, if the MFE lens 108A and the variable aperture 108B are moved integrally in the X' direction within the X'Y' plane by the micro-motion mechanism 108D, which serves as the second telecentric adjustment mechanism as shown in FIG30, then the opening of the aperture 108B (the elliptical region APh' in FIG31) is eccentric relative to the optical axis AXc. Consequently, the surface light source formed within the elliptical region APh' also moves as a whole in the X' direction. In this case, the principal ray (central ray) of the illumination light ILm illuminating the DMD10 can also be tilted in the X'Z plane relative to the optical axis AXb in FIG31, that is, the incident angle θα of the illumination light ILm towards the DMD10 can be changed in the X'Z plane from the initial set angle (35.0°). Furthermore, even if it is configured such that only the aperture 108B moves slightly within the X'Y' plane by means of the micro-motion mechanism 108D, the incident angle θα can be changed in the same way.

[0116] Thus, in order to make the MFE lens 108A and the aperture 108B move relatively large as a unit, it is necessary to pre-enlarge the beam width (diameter of the illumination range) of the illumination light ILm illuminating the MFE lens 108A from the input lens system 104. Furthermore, it is also effective to provide a moving mechanism that moves the illumination light ILm illuminating the MFE lens 108A laterally in the X'Y' plane in conjunction with the amount of this displacement. This moving mechanism can be composed of a mechanism that tilts the orientation of the exit end of the fiber bundle FBn, or a mechanism that tilts the parallel plane plate (quartz plate) disposed near the MFE lens 108A.

[0117] Both the first telecentric adjustment mechanism (drive unit 100C, etc.) and the second telecentric adjustment mechanism (micro-motion mechanism 108D, etc.) can adjust the incident angle θα of the illumination light ILm toward the DMD10. However, regarding the adjustment amount, the first telecentric adjustment mechanism can be used for fine adjustment and the second telecentric adjustment mechanism can be used for coarse adjustment. In actual adjustment, the appropriate choice can be made based on the shape of the pattern to be projected (the amount of telecentric error Δθt or the correction amount), whether to use either the first or second telecentric adjustment mechanism or just one of them.

[0118] Furthermore, the micro-motion mechanism 110C, which serves as the third telecentric adjustment mechanism for offsetting the condenser lens system 110 in the X'Y' plane, has the same effect as the second telecentric adjustment mechanism in offsetting the position of the surface light source defined by the MFE lens 108A and the aperture 108B. However, if the condenser lens system 110 is offset in the X' direction (or Y' direction), the illumination area of ​​the illumination light ILm projected onto the DMD 10 also shifts laterally. Therefore, the degree of this shift is estimated, and the illumination area is set to be larger than the overall size of the mirror surface of the DMD 10. The third telecentric adjustment mechanism using the micro-motion mechanism 110C is also set up for coarse adjustment and used separately from the second telecentric adjustment mechanism.

[0119] [Other Telecentric Adjustment Mechanisms] The adjustment (correction) of telecentric error can also be performed by means of a micro-motion mechanism that moves the position of the exit end of each fiber bundle FBn (n=1~27) shown in Figures 4 and 30 laterally in the X'Y' plane. In this case, the position of the surface light source (a collection of multiple point light sources SPF) formed on the exit surface side of the MFE lens 108A can be finely adjusted in the same way as the first telecentric adjustment mechanism (drive mechanism 100C, etc.) described above.

[0120] The correction of the telecentric error can also be performed by using a micro-motion mechanism such as a micro-head or a piezoelectric actuator to adjust the original angle of the tilting mirror 112 shown in Figures 4, 6, and 31, thereby finely adjusting the incident angle θα (for example, 35.0° in the design) of the illumination light ILm toward the DMD10. Alternatively, the telecentric error can be corrected by finely adjusting the tilt of the mirror surface (central plane Pcc) of the DMD10 using a micro-motion platform that combines the parallel linkage mechanism of the fixed part 10M shown in Figures 4 and 31 with the piezoelectric device. However, the adjustment of the angle of the tilting mirror 112 or the DMD10 is designed for coarse adjustment because the reflected light is tilted at a multiple of the adjustment angle. Furthermore, during the angle adjustment of DMD10, the conjugate plane (optimal focal plane) of the neutral plane Pcc projected onto the substrate P is tilted relative to the plane perpendicular to the optical axis Axa in the direction of scanning exposure (X' direction or X direction).

[0121] When the direction of image plane tilt is the direction of scanning exposure, scanning exposure is performed at the average image plane position of the tilted image plane, so the reduction in contrast of the exposed pattern image is slight. An exposure method that tilts the optimal focal plane of the pattern image projected by the self-projection unit PLU and the surface of the substrate P slightly relative to each other in the direction of scanning exposure, as disclosed in, for example, Japanese Patent No. 2830492, can particularly achieve the effect of increasing the depth of focus (DOF) when exposing isolated patterns.

[0122] Therefore, the function of tilting the DMD10 in the scanning exposure direction (X' direction or X direction) to correct the telecentric error Δθt can also be utilized within a range where the contrast reduction of the pattern image is negligible. When the DMD10 is tilted to a degree that negligible contrast reduction, a certain image plane tilt correction system (such as two wedge-shaped angular prisms) can be provided in the projection unit PLU. Alternatively, a mechanism can be provided to offset a specific lens group or lens in the projection unit PLU relative to the optical axis AXa in order to correct the telecentric error Δθt.

[0123] In the above explanation, it is assumed that the telecentric error Δθt mainly occurs in the X' direction (the tilt direction of the micromirror Msa in the open state). However, the exposed pattern may sometimes be a line and gap pattern on the substrate P that is tilted at angles of 30° to 60° relative to the X' and Y' directions, respectively. In this case, the micromirror Msa in the open state is also tilted in the X'Y' plane and is periodically arranged in a direction orthogonal to its tilt direction. Therefore, the reflected light (imaging beam) Sa' generated from the micromirror Msa in the open state produces a telecentric error Δθt due to the effect of diffraction.

[0124] In the case of the line and gap pattern as shown in Figure 24 above, the centroid error Δθt only occurs in the X' direction. However, in the case of the line and gap pattern that is tilted in the X'Y' plane (within the XY plane), the centroid error Δθt occurs in both the X' and Y' directions. Therefore, even in the case of the line and gap pattern that is tilted at an angle of 30° to 60°, if the possible centroid error Δθt in either the X' or Y' direction exceeds the allowable range, it can be corrected by the centroid error adjustment mechanisms described in Figures 30 and 31 above.

[0125] Furthermore, the telecentric error Δθt is unrelated to the pattern to be projected, and is also caused by the error between the tilt angle θd of the micromirror Ms in the open state of DMD10 and the design value. In the DMD10 illustrated above, the tilt angle θd of the micromirror Msa in the open state is set to the nominal value (design value) 17.5°, but a driving error of ±0.5° occurs due to process deviations during the manufacturing stage of DMD10. When the driving error is ±0.5°, if the incident angle θα of the illumination light ILm onto DMD10 is constant at 35.0°, the telecentric error on the object plane side (DMD10 side) of the projection unit PLU is the largest, reaching ±1°. Therefore, when the projection magnification Mp of the projection unit PLU is 1 / 6, the telecentric error Δθt on the image plane side caused by the driving error of the micromirror Msa is the largest, reaching ±6°.

[0126] Therefore, it is ideal to use the calibration reference unit CU located around the substrate holder 4B in the exposure apparatus shown in Figure 1 above, to measure the telecentric error Δθt caused by the driving error of the DMD10 in each module MUn (n=1~27) in advance, and to adjust (calibrate) it before the actual pattern exposure. An imaging device is assembled in the calibration reference unit CU, which locally magnifies and observes the projected image of the pattern made by the DMD10 obtained by the projection unit PLU. Therefore, by changing the focal position (the position of the substrate holder 4B in the Z direction) little by little, and sampling the projected image of the test pattern (lines and gaps, etc.) made by the DMD10, and analyzing the change in the lateral displacement of the test pattern image at each focal position, the telecentric error Δθt can be measured.

[0127] In the description of the first embodiment above, the so-called isolated pattern as a pattern state is not necessarily limited to the case where a single or a row of micromirrors Ms in the DMD10 are in the open state. For example, the case where 2, 3 (1×3), 4 (2×2), 6 (2×3), 8 (2×4) or 9 (3×3) micromirrors Ms in the open state are densely arranged, and the surrounding micromirrors Ms are, for example, 10 or more micromirrors Msb in the closed state in the X' and Y' directions, can also be regarded as an isolated pattern. Conversely, when 2, 3 (1×3), 4 (2×2), 6 (2×3), 8 (2×4), or 9 (3×3) micromirrors Msb in the closed state are densely arranged, and the micromirrors Ms around them are densely arranged in the X' and Y' directions, for example, several or more (corresponding to several times the size of the isolated pattern) to form a micromirror Msa in the open state, this can also be regarded as a shore-like pattern.

[0128] Furthermore, the line and gap patterns that serve as the pattern are not necessarily limited to the pattern shown in Figure 24, which alternates between one column of open micromirrors Msa and one column of closed micromirrors Msb. For example, it could also be a pattern that alternates between two columns of open micromirrors Msa and two columns of closed micromirrors Msb, a pattern that alternates between three columns of open micromirrors Msa and three columns of closed micromirrors Msb, or a pattern that alternates between two columns of open micromirrors Msa and four columns of closed micromirrors Msb. In any pattern form, if the distribution (density or concentration) of open micromirrors Ms in each unit area (e.g., an area of ​​100×100 micromirrors Ms) of all micromirrors Ms in the DMD10 is known, then the degree of centroid error Δθt or asymmetry of the pattern edge can be easily determined by simulation or the like.

[0129] [Modification 1] In the above embodiment, since the illumination light ILm is tilted towards the DMD10, as shown in FIG11, the intensity distribution of the 0th light equivalent component (light source image Ips) of the imaging beam Sa' formed in the pupil Ep of the projection unit PLU becomes elliptical. Therefore, in order to correct it, as shown in FIG31, an aperture aperture 108B with an opening of an elliptical region APh' is provided on the emission surface side of MFE108A.

[0130] In this situation, as shown in FIG12, the illumination light ILm illuminating the incident surface side of MFE108A becomes a circular illumination area Ef. Therefore, the point light source SPF distributed outside the elliptical area APh' is blocked, resulting in a reduction in the utilization efficiency of the illumination light ILm, i.e., a loss of light. Therefore, in this modified example, the illumination area Ef of the illumination light ILm illuminating the incident surface side of MFE108A is shaped into an ellipse compared to the elliptical area APh'. Therefore, in this modified example, a transmissive diffraction optical element (DOE) is used to minimize the loss of light.

[0131] Figure 33 is a schematic diagram showing the optical configuration of the main parts of the illumination unit ILU using a diffractive optical element (DOE) 200. Figure 33(A) is a diagram showing the configuration of the DOE 200, the input lens system 202 (equivalent to the input lens system 104 in Figure 30 above), and the MFE 108A as viewed in the X'Z plane, and Figure 33(B) is a diagram showing the configuration of the DOE 200, the input lens system 202, and the MFE 108A as viewed in the Y'Z plane. The DOE 200 is formed by engraving fine linear irregularities on the surface of a quartz plate to form a plurality of diffraction grating elements with different spacing or orientation. An illumination optical system that arranges such a DOE 200 near a compound eye lens to change the shape of the surface light source formed on the exit surface side of the compound eye lens is disclosed, for example, in Japanese Patent Application Publication No. 2001-176766.

[0132] In this modified example, the illumination area Ef of the illumination light ILm on the incident surface side of MFE108A is set to be elliptical in the X'Y' plane, and the spacing between the diffraction gratings of DOE200 is set. The illumination light ILm (divergent light) from the fiber bundle FBn shown in Figure 30 above is transformed into a circular parallel beam by a beam expander optical system (not shown) and then incident on DOE200. As shown in Figures 33(A) and 33(B), the diffraction angle of the diffracted light generated from DOE200 is set in such a way that the diffraction angle Δdy when viewed in the Y'Z plane is smaller than the diffraction angle Δdx when viewed in the X'Z plane.

[0133] The position of the back focal point of the input lens system 202 is set at the incident surface of MFE108A. Therefore, due to the difference in diffraction angles Δdx and Δdy, the dimension Uy of the illumination area Ef in the X' direction, which is parallel to the X'Z plane and includes the optical axis AX, is smaller than the dimension Ux of the illumination area Ef in the Y' direction, which is parallel to the Y'Z plane and includes the optical axis AXc. The ratio Uy / Ux, which is the major axis, to the dimension Uy, which is the minor axis, is set to cosθα, as explained above.

[0134] With this configuration, the plurality of point light sources SPF formed on the emission surface side of the MFE108A are formed only on the emission surface side of the plurality of lens devices EL arranged within an elliptical range corresponding to the illumination area Ef. The diffraction grating formed on the DOE200 is such that most (e.g., about 90%) of the total amount of illumination light ILm is incident on the input lens system 202, thus becoming a grating structure that does not produce high-order diffraction light with diffraction angles greater than Δdx and Δdy, or even if it does, it becomes a very low-intensity diffraction light.

[0135] Furthermore, whether using DOE200 or aperture 108B with an elliptical opening, the major and minor axes of the elliptical surface light source formed on the emission surface side of MFE108A are uniquely determined by the incident direction of the illumination light ILm onto DMD10 in the X'Y' plane. Therefore, it is not necessary to adjust the rotation direction of the elliptical region APh' in the X'Y' plane, but if necessary, a small rotation mechanism for DOE200 or aperture 108B can be provided.

[0136] Furthermore, when the incident angle θα of the illumination light ILm toward DMD10 is slightly adjusted from a predetermined value (e.g., 35.0°) in order to adjust (correct) the telecentric error Δθt, the adjustment of the elliptic ratio Uy / Ux of the elliptic region APh' is also considered (refer to Figures 30 and 33). For example, if the incident angle θα is changed by ±0.5° from the predetermined value of 35.0° (equivalent to ±6° in terms of the angle adjustment of the telecentric on the image plane side when the projection magnification Mp = 1 / 6), then the elliptic ratio Uy / Ux at an incident angle of 35.0° becomes 0.8191, the elliptic ratio Uy / Ux at an incident angle of 34.5° becomes 0.8241, and the elliptic ratio Uy / Ux at an incident angle of 35.5° becomes 0.8141. Therefore, the change in the elliptic ratio Uy / Ux caused by an adjustment of ±0.5° in the incident angle θα is negligible, at ±0.6%. Furthermore, even if the adjustment range of the incident angle θα is assumed to be ±1.0°, the change in the elliptic ratio Uy / Ux is still ±1.2%, so there is no problem even if the elliptic ratio Uy / Ux is not adjusted.

[0137] [Modification Example 2] Figure 34 shows the optical configuration of a second modification example in which the distribution (profile) of the illumination light ILm incident on the MFE108A is set to be elliptical in order to form an elliptical surface light source (an assembly of multiple point light sources SPF) on the emission surface side of the MFE108A. Figure 34(A) is a view in the X'Z plane of the arrangement of two cylindrical lenses 210 and 212 positioned near the front of the MFE108A, and Figure 34(B) is a view in the Y'Z plane of the arrangement of Figure 34(A). In this modification example, a cylindrical lens 210 with positive refractive power and a cylindrical lens 212 with negative refractive power are arranged at predetermined intervals along the optical axis AXc of the illumination unit ILu near the front of the MFE108A.

[0138] The generatrices of each cylindrical lens 210 and 212 are set to be parallel to the X' axis. Therefore, if an illumination light ILm with a circular intensity distribution in the cross-section and set as a parallel beam is incident on the cylindrical lens 210, the cylindrical lens 210 functions as a parallel plate without refractive power in the X'Z plane. Therefore, as shown in Figure 34(A), the width of the illumination light ILm in the X' direction does not change and it is directly incident on the cylindrical lens 212. On the other hand, as shown in Figure 34(B), the cylindrical lens 210 has positive refractive power in the Y'Z plane. Therefore, the illumination light ILm passing through the cylindrical lens 210 gradually decreases in width in the X' direction while being incident on the cylindrical lens 212.

[0139] The cylindrical lens 212 functions as a parallel plate without refractive power in the X'Z plane. Therefore, as shown in Figure 34(A), the width of the illumination light ILm in the X' direction remains unchanged, and it is emitted directly as a parallel beam, reaching the illumination area Ef on the MFE108A in the rear section. On the other hand, as shown in Figure 34(B), the cylindrical lens 212 has negative refractive power in the Y'Z plane. Therefore, the illumination light ILm passing through the cylindrical lens 212 reaches the illumination area Ef on the MFE108A in the rear section as a parallel beam with a reduced width in the Y' direction compared to the original width.

[0140] Thus, when two cylindrical lenses 210 and 212 are used as lens devices with anisotropic refractive power, the focal position of cylindrical lens 210 (the position where the illumination light ILm is focused in the Y'Z plane) and the focal position of cylindrical lens 212 are set to be at the same position on the optical axis AXc. By appropriately setting the focal distance between each cylindrical lens 210 and 212 and the interval between cylindrical lenses 210 and 212, the ellipticity ratio Uy / Ux of the illumination area Ef and the incident angle θα can be set to corresponding values ​​(cosθα).

[0141] Furthermore, in the configuration of Figure 34, the circular intensity distribution within the cross-section of the original illumination light ILm is compressed and deformed into an elliptical shape in the Y' direction. However, the arrangement of the cylindrical lenses 210 and 212 can also be changed. The negative cylindrical lens 212 magnifies the illumination light ILm with a circular intensity distribution in the Y' direction, and the positive cylindrical lens 210 transforms it into a parallel beam. In this case, the circular intensity distribution within the cross-section of the original illumination light ILm is transformed into an elliptical shape with the width in the X' direction remaining unchanged and the beam lengthened in the Y' direction. Therefore, when the illumination light ILm passes from the negative cylindrical lens 212 to the positive cylindrical lens 210, the generatrices of each cylindrical lens 210 and 212 are set to be parallel to the Y' axis.

[0142] As described above, the intensity distribution within the cross-section of the illumination light ILm illuminating the MFE108A can be elliptical. This can be achieved by using lens systems with different magnifications in the longitudinal and lateral directions, such as anamorphic lenses, lens devices or reflective devices with annular surfaces, Fresnel lens devices, or microprism arrays, either individually or in combination with other spherical or aspherical lenses. In this case, the anamorphic lens, the lens device with annular surface, the Fresnel lens device, and the microprism array all function as optical devices (lens devices) with anisotropic refractive power or refraction angle, or as distribution-shaping optical devices that shape the contour of the intensity distribution within the cross-section, similar to cylindrical lenses.

[0143] [Modification 3] Regarding projection exposure apparatuses that expose patterns of electronic components such as semiconductor circuits or display panels on semiconductor wafers or glass substrates using a mask substrate, it is known that a deformable illumination method is used to faithfully expose patterns with finer linewidths. This deformable illumination method not only sets the shape of the surface light source formed on the pupil surface (e.g., the exit surface side of a compound eye lens) within an illumination optical system that illuminates the mask substrate to the illumination light, but also sets it to a ring shape, a two-pole shape, or a four-pole shape. Japanese Patent Application Publication No. 2001-176766, listed above, also discloses a configuration using diffractive optical devices for four-pole illumination or ring illumination.

[0144] In the projection exposure of the pattern using DMD10, in order to cope with high resolution, the same configuration corresponding to the anamorphic illumination method can be set in the illumination unit ILU. The simple configuration for the anamorphic illumination method is to set the opening shape of the aperture 108B shown in Figure 31 above to be annular or multipole. Figure 35 is a diagram of a plurality of point light source SPFs (area light source) formed on the exit surface side of MFE108A, viewed in the +Z direction, similar to Figure 31. In the aperture 108B, light from the point light source SPF in the area between the outer ring APro and the inner ring APri, which are only elliptical in outline, is allowed to pass through, while light from the point light source SPF located in the central part inside the inner ring APri, which includes the optical axis AXc, and the peripheral part outside the outer ring APro, is blocked, thus forming an elliptical annular opening.

[0145] In this modified example, the ratio Uy / Ux of the major axis Ux in the X' direction and the minor axis Uy in the Y' direction of the elliptical outer ring APro is also set to cosθα by the incident angle θα. Furthermore, the ratio of the major axis in the X' direction and the minor axis in the Y' direction of the elliptical inner ring APri is also set to cosθα by the incident angle θα. When the aperture 108B is set as an elliptical ring with an opening, it is sufficient to form a light-shielding film obtained from a chromium layer on the entire surface of the transparent quartz plate, and then remove a portion of the light-shielding film in the shape of an elliptical ring with an opening by etching.

[0146] Furthermore, when performing four-electrode illumination as multi-electrode illumination, as shown in FIG36, an aperture aperture 108B is provided with the following opening: Light from the point light source SPF passes through four fan-shaped regions APa1, APa2, APa3, and APa4, each inclined at approximately 45° from the X' axis and Y' axis, within the elliptical annular region surrounded by the outer ring APro and the inner ring APri in FIG35. The shape of the annular region containing all four fan-shaped regions APa1, APa2, APa3, and APa4, or the annular line passing through the center (centroid) point of each fan-shaped region APa1, APa2, APa3, and APa4, is set to an elliptical shape with a ratio having the cosine value (=cosθα) of the incident angle θα.

[0147] In this modified example, an elliptical annular opening or four fan-shaped openings are provided in the aperture 108B for annular or multi-electrode illumination. However, as disclosed in Japanese Patent Application Publication No. 2001-176766 listed above, an elliptical annular illumination area or a quadrupole illumination area can be formed on the incident surface of the MFE108A by using a diffractive optical device (DOE) or a deflecting mirror device (a conical prism with the optical axis as the center of rotation), which can reduce the light loss caused by the light from the point light source SPF being blocked by the light-shielding part of the aperture 108B. Furthermore, the shape of the surface light source of each pole used for multi-electrode illumination is not limited to the fan-shaped areas APa1, APa2, APa3, and APa4 as shown in Figure 36, and can also be a rectangular area or a circular area. In addition to 4 poles, the number of poles can also be 2, 6, or 8 poles.

[0148] [Modification Example 4] Due to the oblique illumination of the illumination light ILm toward DMD10, the distribution of the 0th order light equivalent component of the imaging beam Sa' formed in the pupil Ep of the projection unit PLU (the 9th order diffraction light Id9 in Figure 22 and the 18th order diffraction light Id18 in Figure 26) is deformed into an elliptical shape. This can be corrected to a circular shape by deforming the shape of the surface light source in the illumination unit ILU. However, as illustrated in Figures 23 or 27 above, depending on the settings of the diameter (numerical aperture NA) of the pupil Ep of the projection unit PLU, the intensity distribution Hpa of the 0th order light equivalent component after circular correction (the diameter of the 9th order diffraction Id9 in Figure 23 and the 18th order diffraction Id18 in Figure 27), the arrangement spacing Pdx and Pdy of the micromirrors Ms of the DMD10, the incident angle θα (=2θd), and the wavelength λ of the illumination light ILm, sometimes excess diffraction light components pass through the pupil Ep in large quantities, causing the quality of the pattern image exposed on the substrate P to deteriorate.

[0149] The excess diffracted light component is mainly determined by the spacing Pdx, Pdy between the micromirrors Ms of DMD10 and the wavelength λ. For example, as explained in Figures 22 and 23 above, on both sides of the X' direction (and Y' direction) of the 9th diffracted light Id9, which is the equivalent component of the 0th order light, the 8th diffracted light Id8 and the 10th diffracted light Id10, which are equivalent components of the ±1st order light generated by the spacing Pdx (=Pdy), respectively produce circular intensity distributions Hpb (and Hpc, Hpd). In the case of Figure 23, if the shift ΔDx (telecentric error Δθt) of the intensity distribution Hpa of the 9th diffracted light Id9, which is the equivalent component of the 0th order light, is corrected to zero, then a portion of the intensity distributions Hpb obtained by the 8th diffracted light Id8 and the 10th diffracted light Id10 appear in the pupil Ep of the projection unit PLU.

[0150] When the maximum numerical aperture on the image plane side corresponding to the diameter of the pupil Ep of the projection unit PLU is set to NAI, when the numerical aperture NAI is greater than 0.3 as illustrated in FIG23, the intensity distributions Hpb, Hpc, and Hpd of the ±1 light equivalent components from the DMD10 pass through the pupil Ep in large quantities. Therefore, on the substrate P, although the edge lines of each micromirror Msa in the open state of the DMD10 are weak, they are also resolved, resulting in the degradation of the image quality of the finally exposed pattern.

[0151] Therefore, based on the relationship R = k(λ / NAi), which is determined by the necessary resolution R (μm), the center wavelength λ of the illumination light ILm, and the process factor k (o < k ≦ 1), the numerical aperture NAi on the image plane side corresponding to the maximum effective diameter of the pupil Ep of the projection unit PLU is set. As an example, when the process factor k is set to 0.6, the center wavelength λ is set to 355.00 nm, and the necessary resolution R is set to 0.8 μm, the numerical aperture NAi becomes approximately 0.266. Furthermore, when the center wavelength λ is set to 343.33 nm, with other conditions remaining the same, the numerical aperture NAi becomes approximately 0.257. Furthermore, since the distances between the center wavelength λ and the micromirrors Ms of DMD10, Pdx and Pdy, and the incident angle θα of the illumination light ILm are fixed, the angles on the object side of the 9th diffracted light Id8 and 10th diffracted light Id10, which are equivalent to the 0th diffracted light Id9 and the 0th diffracted light Id9 respectively, after correcting the telecentric error Δθt, are uniquely determined by the above formula (2) or formula (3).

[0152] When the center wavelength λ is set to 355.00 nm, as shown in Figures 22 and 23 above, the diffraction angle θ9 of the 9th diffracted light Id9 is -1.04°, the diffraction angle θ8 of the 8th diffracted light Id8 is +2.73°, and the diffraction angle θ10 of the 10th diffracted light Id10 is -4.81°. Therefore, the angle Δθ8 of the 8th diffracted light Id8 as viewed from the 9th diffracted light Id9 is 3.77°, and the angle Δθ10 of the 10th diffracted light Id10 as viewed from the 9th diffracted light Id9 is also 3.77°. Since the angles Δθ8 and Δθ10 are the same, if we assume Δθ8 = Δθ10 = Δθj and convert it to the numerical aperture NAoj on the object side, then according to NAoj = sin(Δθj), NAoj ≈ 0.06575. When the projection magnification Mp of the projection unit PLU is set to 1 / 6, the numerical aperture NAij on the image side corresponding to the numerical aperture NAoj on the object side becomes NAij = NAoj / Mp ≈ 0.395.

[0153] Furthermore, within the pupil Ep of the projection unit PLU, as shown in Figure 23, an intensity distribution Hpa of circular shape is formed, consisting of an ellipticized, corrected 9th-order diffracted light component Id9. As explained in Figure 11 above, the ratio of the diameter of the intensity distribution Hpa to the diameter of the pupil Ep, i.e., the value σ (0 < σ ≦ 1), is determined by the magnification of the imaging optical path from the surface light source formed on the exit surface side of the MFE108A to the pupil Ep of the projection unit PLU, as shown in Figure 31, and the size of the surface light source formed on the exit surface side of the MFE108A (the size of its opening when the aperture 108B is provided). Therefore, the numerical aperture NAib on the image plane side corresponding to the radius of the intensity distribution Hpa becomes NAib = σ·NAi. Similarly, the intensity distributions Hpb (and Hpc, Hpd) obtained from the 8th diffracted light Id8 and the 10th diffracted light Id10, which are equivalent to ±1st diffracted light components, also have a diameter distribution that is approximately the same as that of the intensity distribution Hpa.

[0154] If the above state is schematically represented, it becomes as shown in Figure 37. In Figure 37, the horizontal line represents the numerical aperture in the X' direction within the plane of the pupil Ep, the vertical line represents the numerical aperture in the Y' direction within the plane of the pupil Ep, and the center point represents the center of the pupil Ep (optical axis AXp). Here, the arrangement spacing Pdx (Pdy) of the micromirrors Ms is set to 5.4 μm, the center wavelength λ is set to 355.00 nm, the maximum numerical aperture NAi on the image plane side of the projection unit Plu is set to 0.266, and the σ value is set to 0.8. Therefore, the numerical aperture NAi on the image plane side of each intensity distribution Hpa, Hpb, Hpc, and Hpd becomes approximately 0.213.

[0155] Under the above conditions, the center (centroid) of the intensity distributions Hpb, Hpc, and Hpd obtained from the ±1st order diffracted light Id8 and Id10, respectively, is large due to the numerical aperture NAij = 0.395, and therefore located outside the pupil Ep, although a portion of them pass through the pupil Ep. Therefore, it is possible to image even though the edge line of the micromirror Msa in the open state is weak. Therefore, it is conceivable to set the intensity distributions Hpb, Hpc, and Hpd so that they are all approximately completely outside the pupil Ep. If the center wavelength λ, the distance between micromirrors Ms Pdx and Pdy, the incident angle θα, and the maximum numerical aperture NAi (NAi = 0.266 here) on the image plane side of the projection unit Plu are not changed, then by adjusting the σ value, the diameter of the intensity distribution Hpa of the 9th diffracted light Id9, which is equivalent to the 0th order light component, i.e. the numerical aperture NAib, is reduced from the state in Figure 37, and the intensity distributions Hpb, Hpc, and Hpd can be arranged almost completely outside the pupil Ep.

[0156] As a specific example, the numerical aperture NAij (≈0.395) of the center (centroid) of the 8th diffracted light Id8 and the 10th diffracted light Id10, which are equivalent to the ±1st diffracted light components, the maximum numerical aperture NAi (≈0.266) of the image plane side of the projection unit PLU, and the numerical aperture NAib of the diameter of the intensity distribution Hpa of the 9th diffracted light Id9, which is equivalent to the 0th diffracted light component, set by the σ value, can theoretically be set to the relationship of the following formula (8).

[0157] [Number 8] NAib≦NAij-NAi (where NAib>0)···(8)

[0158] As explained in the above embodiment, after correcting the ellipticization of the intensity distribution Hpa of the 9th diffracted light Id9, which is equivalent to the 0th order light component, and with the resulting telecentric error Δθt also corrected to zero, the numerical aperture NAib, which is equivalent to the diameter of the intensity distribution Hpa, ideally satisfies Equation (8) and is maximized as much as possible. Therefore, under the conditions of Figure 37, it is ideal to set NAib≈0.129, which, if expressed in terms of σ, becomes σ​=NAib / NAi=0.485. In practical settings, there is an acceptable range, as long as it is within the range of σ=0.485±15%. This setting can also be applied in the annular illumination method or multi-electrode illumination method described in Figures 35 and 36.

[0159] Furthermore, in the case of the line and gap pattern described in Figures 24 to 27 above, the distance between the line and the gap becomes 2Pdx (or 2Pdy). Therefore, the angle subtended by the diffracted light equivalent to the ±1st order light relative to the diffracted light equivalent to the 0th order light is about half that in the case of Figure 37, which is approximately NAij / 2 when converted to numerical aperture. Thus, a large amount of the diffracted light equivalent to the ±1st order light passes through the pupil Ep, and the line and gap pattern is faithfully exposed.

[0160] Figure 38 is a schematic diagram showing the distribution of diffracted light under the conditions illustrated in Figure 37, where the wavelength λ of the illumination light ILm is changed to 343.44 nm and the maximum numerical aperture NAi on the image plane side of the projection unit PLU is set to 0.25. When the arrangement spacing Pdx (=Pdy) of the micromirrors Ms of DMD10 is 5.4 μm, the telecentric error of the 9th order diffracted light Id9 with the smallest tilt angle from the optical axis Axa of the projection unit PLU on the object plane side (DMD10 side) is approximately 0.078° in angle terms, and the telecentric error Δθt on the image plane side is approximately 0.467°. When this value is not acceptable, it is corrected by the telecentric error adjustment mechanism described above.

[0161] Furthermore, in Figure 38, assuming that the intensity distribution of the 9th diffracted light Id9 (the 0th diffracted component) at the pupil Ep is corrected from an elliptical shape to a circular distribution Hpa based on the incident angle θα of the illumination light ILm after correcting the telecentric error Δθt. When the wavelength λ is 343.33 nm, if the center (centroid) of the +1 diffracted light Id8 (the +1 diffracted component) and the center (centroid) of the -1 diffracted light Id10 (the -1 diffracted component) corresponding to the distance Pdx (Pdy) between the micromirror Ms are represented by the numerical aperture NAij, then NAij≈0.382 in the X' and Y' directions, respectively.

[0162] Therefore, based on the above formula (8), when the maximum numerical aperture NAi of the pupil Ep is 0.25, the numerical aperture NAib of the circular intensity distribution Hpa of the 0th order light equivalent component (9th order diffraction Id9) from DMD10 becomes approximately 0.132. Therefore, if the size of the surface light source generated by MFE108A or aperture 108B (specifically, the size of the elliptical minor axis) is set to approximately 0.528 or less in terms of σ (=NAib / NAi), then as shown in FIG38, the intensity distribution Hpb of the excess +1 order light equivalent component 8th order diffraction Id8 or -1 order light equivalent component 10th order diffraction Id10 can be set outside the pupil Ep.

[0163] By setting this σ value, in the case of a line and gap pattern made in the X' direction with a spacing of 2Pdx as shown in Figures 24 and 25, i.e., a line and gap pattern with the smallest grating spacing on the DMD10, the center (centroid) points of the intensity distribution ±Hpb' of the higher-order diffracted light equivalent to the ±1st order light generated by the micromirrors Msa arranged alternately in the X' direction with a spacing of 2Pdx become approximately 0.19 when converted to numerical aperture NA$1. This is equivalent to the numerical aperture NAib of the radius of each intensity distribution ±Hpb' and the numerical aperture NAib of the intensity distribution Hpa. Therefore, within the maximum numerical aperture NAi (=0.25) of the projection unit PLU, an area containing more than half (calculated as approximately 72.6%) of the center (centroid) points of each intensity distribution ±Hpb' appears symmetrically with the optical axis AXa, and the line and gap pattern is well imaged.

[0164] Furthermore, in the case of a line and gap pattern made in the Y' direction with a spacing of 2Pdy, the center (centroid) of the intensity distribution of the higher-order diffracted light of the ±1st order light component ±Hpb' appears on the Y' axis after a light rotation of about 90° around axis Axa from the state in Figure 38.

[0165] Based on the above, it is also considered that the intensity distributions ±Hpb' of the higher-order diffracted light, which are equivalent to the ±1st order light components, are each completely contained within the pupil Ep. In this case, it is sufficient to set the numerical aperture corresponding to the radius of each intensity distribution ±Hpb' shown in Figure 38 as NAI (0.25) - NA$1 (0.19) = 0.06. This means that the numerical aperture NAib of the intensity distribution Hpa is set to approximately 0.06, that is, the σ value is set to approximately 0.24 (=0.06 / 0.25).

[0166] Here, considering the case where the micromirror Msa in the open state is spaced at a distance of Pdx (Pdy), the numerical aperture NAij (≈0.382) of the intensity distribution Hpb of the 8th diffracted light Id8 and 10th diffracted light Id10 of the equivalent ±1st order light, and the case where the micromirror Msa in the open state is spaced at a distance of 2Pdx (2Pdy), the numerical aperture NA$1 (≈0.19) of the intensity distribution ±Hpb' of the equivalent ±1st order light, the intensity distribution ±Hpb' of the equivalent ±1st order light, can be further set as the condition of equation (8) above.

[0167] [Number 9] (NAi-NA$1)≦NAib≦ (NAij-NAi)···(9)

[0168] If the equation (9) is divided by the maximum numerical aperture NAi of the image plane side of the projection unit PLU, it can be transformed into equation (10) which represents the better range of σ value.

[0169] [Number 10] (1-NA$1 / NAi)≦σ≦ (NAij / NAi-1)···(10)

[0170] In the above embodiments or variations, the ellipticity of the light source image Ips formed in the pupil Ep of the projection unit PLU, i.e., the intensity distribution Hpa of the 9th order diffracted light (or 18th order diffracted light, etc.) which is the equivalent component of the 0th order light from DMD10, is suppressed and corrected to a circular shape. Therefore, the intensity distributions Hpb, Hpc, Hpd, and ±Hpb' of the ±1st order light equivalent components also become circular. However, in reality, the illumination light ILm is illuminated at an angle from the X' direction, so even if the intensity distribution Hpa becomes a perfect circle, each intensity distribution Hpb, Hpc, Hpd, and ±Hpb' becomes an ellipse slightly compressed in the X' direction according to the diffraction angle from the center of the intensity distribution Hpa (the position of the optical axis AXa).

[0171] Therefore, the numerical aperture in the X' direction on the image plane side, corresponding to the radius of the intensity distributions Hpb (and Hpc, Hpd) shown in Figures 37 and 38, is strictly speaking a numerical aperture NAib' that is slightly smaller than the numerical aperture NAib of the intensity distribution Hpb. Here, the numerical aperture NAib on the image plane side, corresponding to the radius of the intensity distribution Hpa, becomes a projection magnification Mp (=1 / 6) on the object plane side (DMD10 side), which is Mp·NAib. Therefore, the angle corresponding to the numerical aperture (Mp·NAib) is set as Δθσ (=arcsin(±Mp·NAib)). Similarly, the numerical aperture NAib' in the X' direction on the image plane side of the intensity distributions Hpb (and Hpc, Hpd) becomes a projection magnification Mp (=1 / 6) on the object plane side (DMD10 side), which is Mp·NAib'. Therefore, the angle corresponding to the numerical aperture (Mp·NAib') in the X' direction is set as Δθib (=arcsin(±Mp·NAib')).

[0172] If we consider the angles Δθσ and Δθib, then the above equation (2) or equation (3) is expressed as follows: equation (11) and equation (12) respectively.

[0173] [Number 11] sin(θj+Δθib)=j(λ / Pdx)-sin(θα+Δθσ)···(11)

[0174] [Number 12] sin(θj+Δθib)=sin(θα+Δθσ)-j(λ / Pdx)···(12)

[0175] [Modification 5] Furthermore, even for an exposure apparatus in which the illumination light ILm is incident on the DMD10, the setting of the intensity distribution Hpb of the diffracted light equivalent to the ±1st order light generated by the arrangement spacing Pdx and Pdy of the micromirrors Ms of the DMD10 can be similarly applied to the setting of the σ value outside the pupil Ep of the projection unit PLU, as shown in Figures 37 and 38. In this case, a polarizing beam splitter and a 1 / 4 wavelength plate are provided in the optical path between the DMD10 and the projection unit PLU. In this case, as an example, the polarization separation surface of the polarizing beam splitter is set at 45° with the optical axis AXa of the projection unit PLU, and the illumination light ILm is set approximately perpendicular to the central surface Pcc of the DMD10 from the polarizing beam splitter side (incident angle θα≈0°).

[0176] In this configuration, the optical axes (AXc, AXb) of the condenser lens system emitting the illumination light ILm and the optical axis Axa of the projection unit PLU are coaxial or parallel on the DMD10. Therefore, there is no ellipticity problem, but a telecentric error Δθt is generated due to the driving error of the micromirror Msa in the open state of the DMD10, which needs to be corrected. Furthermore, in the incident illumination mode, when the optical axes of the condenser lens system and the optical axis Axa of the projection unit PLU are not strictly parallel on the DMD10, but are deliberately tilted at a certain incident angle θα (e.g., 5° or more), it is more ideal to perform ellipticity correction.

[0177] [Modification 6] In the elliptic correction method described in the above embodiments or modifications, the outline of the light source image Ips (the intensity distribution Hpa of the diffracted light equivalent to the 0th order light component from DMD10) formed as an ellipse in the pupil Ep of the projection unit PLU is corrected to a circle, and the outline shape of the surface light source generated in the illumination unit ILU is set to a complementary ellipse. The incident angle θα of the illumination light ILm towards DMD10 does not change significantly in the device configuration, so the ratio of the ellipse can be considered to be approximately constant.

[0178] Therefore, the outline shape of the surface light source generated in the illumination unit ILU can also be kept circular. At a position near or in front of the pupil Ep in the projection unit PLU, cylindrical lenses, toroidal lenses, diffraction optics, and other distribution shaping optical devices with non-isotropic refractive power can be added to correct the distribution of the elliptical light source image Ips (the intensity distribution Hpa of the diffracted light equivalent to the 0th order light component from DMD10) to be circular. However, in the case of this projection unit PLU, the optical design of the imaging lens including the distribution shaping optical device needs to be carried out in a way that allows the imaging performance of the projection unit PLU, especially the isotropicity, distortion characteristics, and other aberration characteristics of the projection magnification Mp, to be within an acceptable range.

[0179] The pattern exposure apparatus EX described in the above embodiments or variations includes: a DMD10 (spatial light modulation device) having a plurality of micromirrors Ms that are selectively driven based on the drawing data; an illumination unit PLU that illuminates the DMD10 with illumination light ILm at a predetermined incident angle θα; and a projection unit PLU that receives reflected light from the selected on state of the micromirrors Msa of the DMD10 and projects it onto the substrate P.

[0180] The illumination unit ILU includes a condenser lens system 110, which serves as a condenser optical component. This system focuses and obliquely illuminates the DMD 10 with light from a surface light source (an assembly of multiple point light source SPFs on the exit surface side of the MFE108A) of a predetermined shape, which is the source of the illumination light ILM. The system is arranged along optical axes (AXb, AXc) that are oblique to the optical axis AXa of the projection unit PLU at an incident angle θα, thereby optically conjugating the surface light source (the assembly of point light source SPFs) with the pupil Ep of the projection unit PLU. This serves as a corrective optical component. The aperture 108B of the device corrects the shape deformation of the outline of the surface light source Ips (intensity distribution Hpa) in the following ways: the overall outline of the intensity distribution Hpa of the surface light source image Ips formed on the pupil Ep of the projection unit PLU by diffracted light (Id9, Id18, etc., which are equivalent to the zero-order light components of the reflected light from the micromirror Msa in the open state of DMD10) is deformed into an elliptical shape according to the incident angle θα; and the diffractive optical device (DOE) 200 or lens devices 210, 212 with anisotropic refractive power, etc.

[0181] Furthermore, according to the above embodiments or variations, in a device manufacturing method that uses illumination light ILm from illumination unit ILU to illuminate a DMD10 (spatial light modulator) having a plurality of micromirrors Ms that are selectively driven to depict patterns based on electronic components (semiconductor circuits, displays, wiring, sensors, etc.), and projects the reflected light from the selected on-state micromirrors Msa from DMD10 onto substrate P via projection unit ILm, thereby forming a pattern of electronic components on substrate P, the following applies: The following stages are performed: light from a surface light source of a predetermined shape generated within the illumination unit ILU is focused as illumination light and obliquely irradiated onto the DMD10 at an incident angle θα; and the shape of the surface light source's outline is deformed in a manner that corrects the following situation, namely, the outline of the light source image Ips (intensity distribution Hpa) of the surface light source formed on the pupil Ep of the projection unit PLU by the reflected light (diffraction light Id9, Id18 of the 0th order light equivalent component) from the micromirror Msa in the open state of the DMD10 is deformed into an elliptical shape according to the incident angle θα.

[0182] According to this configuration, the overall outline of the intensity distribution Hpa of the light source image Ips formed on the pupil Ep of the projection unit PLU, that is, the 0th order light equivalent component of the micromirror Msa in the open state of DMD10, can be corrected from a deformed ellipse to a circular shape, so that the image quality of the edge of the pattern image projected and exposed by the projection unit PLU on the substrate P is uniform regardless of the directionality of the edge. [Simplified Explanation of the Diagram]

[0009] [Fig. 1] is a perspective view showing the general appearance of the pattern exposure apparatus EX of this embodiment. [Fig. 2] is a diagram showing an example of the arrangement of the projection areas IAn of DMD10 projected onto the substrate P by the projection units PLU of each of the plurality of exposure module groups MU. [Fig. 3] is a diagram illustrating the state of successive exposures performed using the four specific projection areas IA8, IA9, IA10, and IA27 in Fig. 2. [Fig. 4] is an optical configuration diagram showing the specific structure of the two exposure modules MU18 and MU19 arranged in the X direction (scanning exposure direction) as observed in the XZ plane. [Fig. 5] is a diagram schematically showing the state of the tilt angle θk between DMD10 and the illumination unit PLU in the XY plane. [Fig. 6] is a diagram detailing the imaging state of the micromirror of DMD10 obtained by the projection unit PLU. [Fig. 7] is a schematic diagram of the MFE lens 108A, which serves as an optical integrator 108, viewed from the exit surface side. [Figure 8] is a schematic diagram showing an example of the configuration relationship between the point light source SPF formed on the exit surface side of the lens device EL of the MFE lens 108A in Figure 7 and the exit end of the fiber bundle FBn. [Figure 9] is a schematic diagram showing the state of the light source image formed on the pupil Ep of the projection unit PLU shown in Figure 6. [Figure 10] is a simplified optical path diagram of Figure 6. [Figure 11] is a schematic diagram showing the state of the light source image Ips formed on the pupil Ep by the zero-order light equivalent component of the imaging beam Sa from the DMD 10. [Figure 12] is a schematic diagram of the elliptical light source surface of the MFE lens 108A of the optical integrator 108 viewed from the exit surface side, similar to Figure 7. [Figure 13] is a schematic diagram showing the behavior of the imaging beam Sa in the optical path from the pupil Ep of the projection unit PLU shown in Figure 6 to the substrate P. [Figure 14] is a magnified perspective view of the state of a portion of the micromirrors Ms of the DMD10 when the power supply to the drive circuit of the DMD10 is disconnected. [Figure 15] is a magnified perspective view of a portion of the mirror surface of the DMD10 with the micromirrors Ms in the on and off states. [Figure 16] shows a portion of the mirror surface of the DMD10 viewed in the X'Y' plane, and shows the case where only a row of micromirrors Ms arranged in the Y' direction is in the on state. [Figure 17] is a view of the a-a' arrow portion of the mirror surface of the DMD10 in Figure 16 viewed in the X'Z plane. [Figure 18] is a schematic diagram in the X'Z plane showing the imaging state obtained by the projection unit PLU of the reflected light (imaging beam) Sa from the isolated micromirror Msa as shown in Figure 16. [Figure 19] is a schematic diagram showing the point image intensity distribution Iea of ​​the diffracted image of the normal reflected light Sa from the isolated micromirror Msa at the pupil Ep.[Fig. 20] is a diagram showing a portion of the mirror surface of DMD10 viewed in the X'Y' plane, and also showing the situation where multiple micromirrors Ms adjacent in the X' direction are simultaneously in the open state. [Fig. 21] is a diagram showing the a-a' arrow view of the mirror surface of DMD10 in Fig. 20 viewed in the X'Z plane. [Fig. 22] is a diagram showing an example of the distribution of the angle θj of the diffracted light Idj generated by DMD10 in the states of Fig. 20 and Fig. 21. [Fig. 23] is a diagram schematically showing the intensity distribution of the imaging beam at the pupil Ep when the diffracted light is generated as in Fig. 22. [Fig. 24] is a diagram showing the state of a portion of the mirror surface of DMD10 when projecting a line and gap pattern in the X'Y' plane. [Fig. 25] is a diagram showing the a-a' arrow view of the mirror surface of DMD10 in Fig. 24 viewed in the X'Z plane. [Fig. 26] is a diagram showing a variation of the distribution of the distribution section of this embodiment. [Fig. 27] is a diagram showing an example of the distribution of the angle θj of the diffracted light Idj generated by the DMD10 in the states of Fig. 24 and 25. [Fig. 28] is a diagram showing the intensity distribution of the imaging beam at the pupil Ep when the diffracted light is generated as in Fig. 26. [Fig. 29] is a diagram showing the contrast result of a spatial image of a line and gap pattern with a linewidth of 1 μm on the image plane. [Fig. 20] is a diagram showing the relationship between wavelength λ and telecentric error Δθt calculated based on equation (2). [Fig. 31] is a diagram showing the specific configuration of the optical path from the fiber bundle FBn in the illumination unit ILU shown in Fig. 4 or 6 to the MFE108A. [Fig. 32] is a diagram showing the specific configuration of the optical path from the MFE108A in the illumination unit ILU shown in Fig. 4 or 6 to the DMD10. [Figure 32] is an exaggerated representation of the state of the point light source SPF formed on the exit surface side of MFE108A when the illumination light ILm incident on MFE108A is tilted in the X'Z plane. [Figure 33] is a diagram showing a modified example of using a diffraction optics (DOE) to set the surface light source formed on the exit surface side of MFE108A as an elliptical illumination unit ILU. [Figure 34] is a diagram showing the optical configuration of a second modified example for forming an elliptical surface light source on the exit surface side of MFE108A. [Figure 35] is a diagram showing an assembly (surface light source) of multiple point light source SPFs formed on the exit surface side of MFE108A viewed from the +Z direction. [Figure 36] is a diagram schematically showing the configuration of the four fan-shaped regions of the surface light source when the elliptical ring opening of Figure 35 is deformed for quadrupole illumination. [Figure 37] is a schematic diagram showing the distribution of diffracted light generated by the arrangement spacing of the micromirrors Ms of the DMD10 on the surface of the pupil Ep of the projection unit PLU. [Figure 38] is a schematic diagram showing the diffracted light appearing on the surface of the pupil Ep of the projection unit PLU when the wavelength λ and the maximum numerical aperture of the pupil Ep as illustrated in Figure 37 are changed.

Claims

1. A pattern exposure apparatus comprising: a spatial light modulator having a plurality of micromirrors selectively driven based on drawing data; an illumination unit illuminating the spatial light modulator with illumination light at a predetermined incident angle; and a projection unit projecting reflected light from the selected on-state micromirrors of the spatial light modulator onto a substrate; wherein the pattern exposure apparatus projects and exposes a pattern corresponding to the drawing data onto the substrate; and the illumination unit comprises: A focusing optical component that focuses light from a surface light source of a predetermined shape, which serves as the source of the illumination light, onto the spatial light modulation device at an angle, and is arranged along an optical axis that is tilted relative to the optical axis of the projection unit at the incident angle, so that the surface light source and the pupil of the projection unit are optically conjugate; and a correcting optical component that corrects the shape deformation of the outline of the surface light source in such a way that the outline of the image of the surface light source formed on the pupil of the projection unit by the reflected light from the micromirror in the open state of the spatial light modulation device is deformed into an elliptical shape according to the incident angle.

2. The pattern exposure apparatus as described in claim 1, wherein, The aforementioned illumination unit includes an optical integrator that generates the aforementioned surface light source by means of an assembly of a plurality of point light sources; the aforementioned focusing optical component is composed of a focusing lens system, which uses illumination light from each of the aforementioned plurality of point light sources to perform Köhler illumination on the aforementioned spatial light modulation device.

3. The pattern exposure apparatus as described in claim 2, wherein, The aforementioned corrective optical component is composed of an aperture, which distorts the shape of the outline of the surface light source by allowing illumination light from a portion of the assembly of the plurality of point light sources to pass through the condenser lens system.

4. The pattern exposure apparatus as described in claim 3, wherein, The aforementioned aperture has an elliptical opening that rotates the major and minor axes of the elliptical outline of the image of the surface light source formed in the pupil of the projection unit by 90°.

5. The pattern exposure apparatus as described in claim 2, wherein, The aforementioned corrective optical component is composed of a distribution shaping optical device, which shapes the distribution of illumination light incident on the aforementioned optical integrator in accordance with the shape of the contour of the aforementioned surface light source to be deformed.

6. The pattern exposure apparatus as claimed in claim 5, wherein, The aforementioned distribution shaping optical device is composed of any one of a diffraction optical device, a cylindrical lens, or a Fresnel lens. The aforementioned diffraction optical device, cylindrical lens, or Fresnel lens sets the overall profile of the distribution of illumination light incident on the aforementioned optical integrator to an ellipse shape in which the major axis and minor axis of the elliptical profile of the image of the aforementioned surface light source formed in the pupil of the aforementioned projection unit are rotated by 90°.

7. The pattern exposure apparatus as described in any of claims 1 to 6, wherein, The outline of the surface light source, which is deformed by the aforementioned modified optical components, is set to be an elliptical shape in which the ratio of the size in the minor axis direction to the size in the major axis direction is the cosine of the incident angle.

8. A method for manufacturing an electronic component, wherein a spatial light modulator that selectively drives a plurality of micromirrors having a pattern for use in an electronic component is tilted and illuminated by illumination light from an illumination unit, and reflected light from the selected on-state micromirrors of the spatial light modulator is projected onto a substrate via a projection unit to form a pattern of the electronic component on the substrate, and the method comprises the following steps: focusing light from a surface light source of a predetermined shape generated in the illumination unit as the illumination light and tilting it onto the spatial light modulator at a predetermined incident angle; and deforming the shape of the outline of the surface light source in a manner that corrects the situation where the outline of the image of the surface light source formed on the pupil of the projection unit by the reflected light from the on-state micromirrors of the spatial light modulator is deformed into an elliptical shape according to the incident angle.

9. An exposure apparatus comprising: a spatial light modulator having a plurality of micromirrors; an illumination unit that illuminates the spatial light modulator with illumination light; and a projection unit that projects reflected light from the micromirrors in an open state of the spatial light modulator; wherein the illumination unit comprises: a first cylindrical lens; and a focusing optical member that focuses light passing through the first cylindrical lens and illuminates the spatial light modulator; the focusing optical member focusing light from a surface light source formed by the illumination light passing freely through the first cylindrical lens and illuminating the spatial light modulator; the first cylindrical lens corrects for the shape distortion of the surface light source distribution in the pupil of the projection unit, whereby the shape of the surface light source distribution is elliptical.

10. The exposure apparatus of claim 9, further comprising: a second cylindrical lens, different from the first cylindrical lens; the second cylindrical lens being located in the optical path between the first cylindrical lens and the condensing optical member; the first cylindrical lens having positive refractive power in a first plane including the optical axis of the illumination unit; and the second cylindrical lens having negative refractive power in the first plane.

11. The exposure apparatus as claimed in claim 10, wherein, The first cylindrical lens described above does not have refractive power in the second plane that is perpendicular to the first plane and includes the optical axis; the second cylindrical lens described above does not have refractive power in the second plane.

12. An exposure apparatus comprising: a spatial light modulator; an illumination unit that illuminates the spatial light modulator with illumination light; and a projection unit that projects the illumination light from the spatial light modulator; the illumination unit illuminating the illumination light with illumination light from a direction different from the optical axis of the projection unit; at least one of the illumination unit and the projection unit comprising an optical member that deforms the illumination light by changing the ratio of the width in a first direction perpendicular to the direction of travel of the illumination light to the width in a second direction orthogonal to the first direction; and at least one of the illumination unit and the projection unit illuminating the deformed illumination light with illumination light from the spatial light modulator.

13. The exposure apparatus as claimed in claim 12, wherein, The optical component included in the above-mentioned illumination unit or projection unit includes an aperture with a non-circular opening.

14. The exposure apparatus as claimed in claim 12, wherein, The optical component included in the above-mentioned lighting unit or the above-mentioned projection unit includes a diffraction optical device in which the diffraction angle in the first direction is different from the diffraction angle in the second direction.

15. The exposure apparatus as claimed in claim 12, wherein, The optical component included in the above-mentioned illumination unit or projection unit includes: a first cylindrical lens having positive refractive power in a first plane including the above-mentioned travel direction; and a second cylindrical lens having negative refractive power in the first plane.

16. The exposure apparatus as claimed in claim 15, wherein, The first cylindrical lens described above does not have refractive power in the second plane that is perpendicular to the first plane and includes the direction of travel; the second cylindrical lens described above does not have refractive power in the second plane.

17. The exposure apparatus as claimed in claim 12, wherein, The angle between the direction of travel of the illumination light illuminating the aforementioned spatial light modulation device and the aforementioned optical axis is 5° or more.

18. An exposure apparatus as described in any of claims 12 to 17, wherein, The aforementioned illumination unit and the aforementioned projection unit are configured such that diffracted light from the aforementioned spatial light modulation device is incident on the aforementioned projection unit at least once; the aforementioned spatial light modulation device includes a plurality of mirrors that are selectively driven based on the depiction data.

19. An exposure apparatus as described in any of claims 12 to 17, wherein, The first direction mentioned above corresponds to a direction perpendicular to the plane containing the travel direction of the illumination light illuminating the spatial light modulation device and the optical axis; the optical component included in the illumination unit or the projection unit deforms the illumination light in such a way that the width of the first direction becomes smaller than the width of the second direction in the cross-section.

20. A method for manufacturing a component, comprising: exposing a substrate using an exposure apparatus as described in any one of claims 9 to 19; and developing the exposed substrate.

Images