JP7567935B2 - Pattern exposure apparatus and pattern exposure method - Google Patents

Description

The present invention relates to a pattern exposure apparatus and a pattern exposure method that expose a pattern of an electronic device or the like onto a substrate using a drawing beam whose intensity is modulated according to drawing data.

Conventionally, in the process of manufacturing fine electronic devices on a substrate, photolithography processing is carried out, including an exposure process in which an exposure beam (light beam, electron beam, etc.) corresponding to the pattern of the electronic device (a pattern that defines the shape of the wiring layer, electrode layer, semiconductor layer, insulating layer, etc.) is irradiated onto a resist layer as a photosensitive layer on the substrate, and a development process in which the substrate after exposure is developed to make a pattern appear from the remaining film part and the removed part of the resist layer. As an exposure device used in the exposure process, a maskless method is known in which the exposure beam is dynamically intensity-modulated based on drawing data (CAD data) corresponding to the pattern to be exposed. As one of the maskless methods, a spot scanning type drawing device is known in which a spot light that is intensity-modulated in response to drawing data is scanned at high speed with a rotating polygon mirror, like a laser beam printer. In the spot scanning method, the intensity of a fine circular spot light projected onto the surface of a photosensitive substrate is usually turned on/off according to digital drawing data (binary information in bitmap format). Therefore, when drawing a fine diagonal line (an edge that is inclined diagonally with respect to the scanning direction of the spot light), stepped jaggedness is likely to occur at the edge of the diagonal line pattern created by the resist layer that appears after development.

In order to reduce such jagged edges, a laser beam printer, as disclosed in Japanese Patent Application Laid-Open No. 5-232414, is known in which a beam from a semiconductor laser modulated according to an image signal (drawing data) is made to pass through a variable aperture element made of an optical microshutter array made of electro-optic crystal before being incident on a rotating polygon mirror, and the beam reflected by the rotating polygon mirror is imaged as a spot on a photoconductor via an imaging lens (fθ lens) or the like. The variable aperture element is configured to change the aperture amount and aperture position by electro-optic modulation (polarization switching) to change the size and center position of the beam. In this way, Japanese Patent Application Laid-Open No. 5-232414 makes it possible to set multiple gradations within one dot by making the beam diameter (spot diameter) in the sub-scanning direction on the photoconductor surface variable, and by making the center position of the beam variable in the sub-scanning direction, the recording pitch by the spot in the sub-scanning direction is changed, and an image output with reduced jagged edges of oblique lines is obtained.

When using a variable aperture element as in JP-A-5-232414, a part of the beam cross section is inevitably blocked, and the light quantity (intensity) of the spot on the photoconductor surface decreases. For this reason, in JP-A-5-232414, a part of the light output from the semiconductor laser is monitored by a light receiving element, and an optical-electrical negative feedback loop is provided to control the driving current of the semiconductor laser so that the light receiving signal and the light emission level command signal are equal, thereby correcting the light quantity fluctuation (decrease). However, if the light output from the semiconductor laser is originally set large in order to minimize the time required for image output to the photoconductor, there is a limit to the increase in the light output of the semiconductor laser that is commensurate with the loss of the beam light quantity due to the variable aperture element. In addition, in JP-A-5-232414, in order to reduce jaggedness, the spots (dots) projected on the photoconductor remain circular, with spots of different diameters connected together, and adjacent spots (dots) are not necessarily connected to each other. Therefore, in the method of varying the spot (dot) diameter as in JP-A-5-232414, when forming a fine electronic pattern, particularly a wiring pattern with a narrow line width, there is a concern that a part of the pattern may be broken.

A first aspect of the present invention is a pattern exposure device having a drawing unit that draws a pattern on a substrate by scanning a spot light generated by a beam supplied from a light source device in a main scanning direction, the device comprising: a first light source device that emits a first beam; a second light source device that emits a second beam; a beam combining section that combines the first beam from the first light source device and the second beam from the second light source device so that they are incident on the drawing unit; a beam shape deformation section that changes the cross-sectional shapes of the first beam and the second beam incident on the beam combining section so that the shape of the first spot light generated by the first beam projected on the substrate and the shape of the second spot light generated by the second beam are changed from each other; and a control device that controls the drawing so that at least an edge portion of a pattern to be drawn on the substrate is drawn with either one or both of the first spot light and the second spot light.

A second aspect of the present invention is a pattern exposure device having a drawing unit that draws a pattern on a substrate by scanning a spot light of a beam supplied from a light source device in a main scanning direction, the device comprising: a light splitting section that splits a beam having a circular cross-sectional shape emitted from the light source device into a first beam and a second beam; a first beam shape deformation section that is provided in the optical path of the first beam and that deforms the cross-sectional shape of the first beam from a circular shape to make the shape of the first spot light on the substrate by projecting the first beam a first shape; a second beam shape deformation section that is provided in the optical path of the second beam and that deforms the cross-sectional shape of the second beam from a circular shape to make the shape of the second spot light on the substrate by projecting the second beam a second shape different from the first shape; a beam combining section that combines the first beam from the first beam shape deformation section and the second beam from the second beam shape deformation section to be incident on the drawing unit; and a control device that controls the pattern to be drawn on the substrate to be drawn with either the first spot light or the second spot light.

A third aspect of the present invention is a pattern exposure method using a drawing unit that draws a pattern on the substrate by scanning spot light from a beam supplied from a light source device in a main scanning direction according to pixel information for each pixel defined in drawing data, and when an edge pixel that forms an edge portion of a pattern that extends diagonally intersecting the main scanning direction is included in the row of pixels scanned in the main scanning direction, the shape of the spot light projected onto at least the edge pixel is set to a slot shape or an oblong shape with its major axis tilted in the direction along the diagonal extension of the pattern.

A fourth aspect of the present invention is a pattern exposure device having a drawing unit that draws a pattern on a substrate by scanning a spot light of a beam supplied from a light source device in a main scanning direction, the device having a beam shape deformation section that receives the beam from the light source device and directs the beam having a deformed cross-sectional shape to the drawing unit, the beam shape deformation section deforming the cross-sectional shape of the beam from a circular shape to a non-circular shape.

1 is a perspective view showing a schematic overall configuration of a pattern exposure apparatus according to a first embodiment; 2 is a perspective view showing a schematic internal configuration of the imaging unit MU1 as a representative of the four imaging units MU1 to MU4 shown in FIG. 1. 3A and 3B are exaggerated views showing the states of three drawing beams B1a, B1b, and B1c passing through the beam expander BEX in the drawing unit MU1 shown in FIG. 2 is a diagram showing a schematic configuration of light source devices LS1A, LS1B, and LS1C and a beam combining unit BD1A shown in FIG. 1. 5A to 5C are diagrams that show schematic arrangements of optical members within the beam shape modifying unit 10B (or 10C) shown in FIG. 2 is a perspective view showing an outline of the arrangement of optical members and optical paths in the beam switching unit BD1B shown in FIG. 1. 7A and 7B are exaggerated views showing the state in which each of the diffracted beams of the three beams LB1a, LB1b, and LB1c incident on the first-stage acousto-optical modulation element AM3 in FIG. 6 is branched toward the corresponding drawing unit MU3. 11 is a perspective view showing states of beams Bna, Bnb, and Bnc incident on each of the writing units MU1 to MU4 from each of the beam switching units BD1B and BD2B. FIG. FIG. 9A shows an example of line-and-space patterns PT1, PT2, PT3 exposed onto a sheet substrate P, and FIG. 9B shows an example of a pixel map (bitmap) on the drawing data of an enlarged portion of the pattern. 9A and 9B. FIG. 9B is a diagram for explaining an operation for drawing a part of one oblique line pattern in the pattern PT2 shown in FIG. 9C is a diagram illustrating bitmap information corresponding to each of pixel rows (data rows) AL1 and AL2 in the drawing data for drawing a part of the diagonal line shown in FIG. 9B. 12A is a diagram showing the optical path within the beam switching unit BD1B (BD2B) of modified example 1, and FIG. 12B is a diagram showing the arrangement of spot lights SPa, SPb, and SPc projected onto the sheet substrate P by the optical path of FIG. 12A. FIG. 5C is a diagram showing a schematic configuration of a modified example of the beam compression system OM2 shown in FIGS. 5A to 5C. FIG. 5 is a diagram showing a configuration of the beam synthesis unit BD1A (BD2A) shown in FIG. 4 according to a second embodiment. 15A and 15B are diagrams illustrating an example of a pattern writing operation using the beam combining unit BD1A of FIG. 14, the beam switching unit BD1B of FIG. 6, and the writing unit MU3 (or MU1). 13A to 13C are diagrams showing an example of a special exposure drawing operation in which the amount of exposure given to each peripheral edge portion of a plurality of rectangular patterns arranged in a matrix is increased; FIG. 13 is a diagram showing a schematic configuration of a beam combining section according to a fourth modified example using only two light source devices. 18A to 18D are diagrams showing schematic combinations of the major axis directions of the spot beams SPb and SPc that can be switched in accordance with the directionality of the edge portion of the pattern using the configuration of FIG. FIG. 13 is a diagram illustrating a modified optical configuration in which two slot-shaped (elliptical) light spots SPb and SPc are produced by a beam LBe from one light source device LSe. 13 is a diagram showing a schematic configuration of a modified example of a light source device and a beam combining unit BD1A (BD2A) including beam shape transformers 10B and 10C. FIG.

Preferred embodiments of the pattern exposure apparatus and pattern exposure method according to the present invention are presented and described in detail below with reference to the accompanying drawings. Note that the present invention is not limited to these embodiments, and includes various modifications or improvements. In other words, the components described below include those that a person skilled in the art would easily imagine and those that are substantially the same, and the components described below can be combined as appropriate. Furthermore, various omissions, substitutions, or modifications of the components can be made without departing from the spirit of the present invention.

First Embodiment
Fig. 1 is a perspective view showing a schematic overall configuration of a pattern exposure apparatus according to a first embodiment. As shown in Fig. 1, the pattern exposure apparatus of this embodiment exposes various patterns corresponding to electronic devices (display devices, wiring devices, sensor devices, etc.) to a photosensitive layer (resist layer) coated on a flexible long sheet substrate P (hereinafter, also simply referred to as substrate P) in a maskless manner by scanning spot light. Such pattern exposure apparatuses are disclosed, for example, in International Publication No. WO 2015/152218, International Publication No. WO 2015/166910, International Publication No. WO 2016/152758, International Publication No. WO 2017/057415, etc.

As shown in Figure 1, the pattern exposure apparatus EX of this embodiment is installed on the floor of the installation location (factory, etc.) parallel to the XY plane of the Cartesian coordinate system XYZ, in which the gravity direction is the Z axis. The exposure apparatus EX includes a rotating drum DR for stably supporting the sheet substrate P and transporting it in the longitudinal direction at a constant speed, four drawing units MU1 to MU4 for drawing a pattern on the photosensitive layer of the sheet substrate P, three light source devices LS1A, LS1B, LS1C (collectively referred to as light source device LS1) for supplying drawing beams B1, B3 to each of the odd-numbered drawing units MU1, MU3, three light source devices LS2A, LS2B, LS2C (collectively referred to as light source device LS2) for supplying drawing beams B2, B4 (not shown in FIG. 1) to each of the even-numbered drawing units MU2, MU4, beam combining units BD1A, BD2A, and beam switching units BD1B, BD2B.

The beam combining unit BD1A combines the beams from the light source devices LS1A, LS1B, and LS1C under predetermined conditions (described in detail below) and sends them to the beam switching unit BD1B, while also transforming the cross-sectional shape of the beams from the light source devices LS1B and LS1C from a circle to a slot shape (long ellipse). Similarly, the beam combining unit BD2A combines the beams from the light source devices LS2A, LS2B, and LS2C under predetermined conditions (described in detail below) and sends them to the beam switching unit BD2B, while also transforming the cross-sectional shape of the beams from the light source devices LS2B and LS2C from a circle to a slot shape (long ellipse). Note that the cross-sectional shape of the beams from the light source devices LS1A and LS2A remains approximately circular. Each of the light source devices LS1 and LS2 is a fiber amplifier laser light source such as those disclosed in WO 2015/166910 and WO 2017/057415, and oscillates an ultraviolet beam having a central wavelength in the ultraviolet wavelength band of 400 nm or less in bursts at several hundred MHz (e.g., 400 MHz) in accordance with drawing data (pixel bit data expressed by the two values "0" and "1").

The beam switching unit BD1B includes a first-stage acousto-optical modulator and a second-stage acousto-optical modulator arranged to pass the three combined beams in series simultaneously or non-simultaneously. The beam diffracted and deflected by the first-stage acousto-optical modulator is sent to the odd-numbered drawing unit MU3, and the beam diffracted and deflected by the second-stage acousto-optical modulator is sent to the odd-numbered drawing unit MU1. Similarly, the beam switching unit BD2B includes a first-stage acousto-optical modulator and a second-stage acousto-optical modulator arranged to pass the three combined beams in series simultaneously or non-simultaneously. The beam diffracted and deflected by the first-stage acousto-optical modulator is sent to the even-numbered drawing unit MU4, and the beam diffracted and deflected by the second-stage acousto-optical modulator is sent to the even-numbered drawing unit MU2.

The rotating drum DR has a cylindrical outer peripheral surface with a constant radius from the rotation center line AXo parallel to the Y axis of the XY plane, and a shaft Sft that protrudes coaxially with the rotation center line AXo to both ends of the rotating drum DR in the Y direction. The sheet substrate P is supported in close contact in the longitudinal direction along the outer peripheral surface of approximately half the circumference of the rotating drum DR, and is transported at a constant speed in the longitudinal direction by the uniform rotation of the rotating drum DR due to the rotation torque from a rotation drive motor (not shown). The base material of the sheet substrate P is a resin material such as a PET (polyethylene terephthalate) film, a PEN (polyethylene naphthalate) film, or a polyimide film, but may also be, for example, a glass material formed into an extremely thin sheet shape with a thickness of 100 μm or less and made flexible, a metal material such as stainless steel formed into a thin sheet shape by rolling, or a paper material containing cellulose nanofibers.

Multiple drawing units MU1 to MU4 are arranged in a line in the Y direction in the space above the rotating drum DR, with each of the odd-numbered drawing units MU1, MU3 and each of the even-numbered drawing units MU2, MU4 being arranged symmetrically with respect to a central plane that is parallel to the YZ plane and includes the rotation center line AXo when viewed in the XZ plane. Each of the odd-numbered drawing units MU1, MU3 is arranged so that the extension of the center line of the beams B1, B3 (beam B3 is not shown in FIG. 1) projected onto the sheet substrate P faces the rotation center line AXo and is tilted by a certain angle (θu) counterclockwise from the central plane when viewed in the XZ plane. Similarly, each of the even-numbered drawing units MU2, MU4 is positioned so that the extension of the center line of the beams B2, B4 (beam B4 is not shown in Figure 1) projected onto the sheet substrate P is directed toward the rotation center line AXo and is tilted a certain angle (θu) clockwise from the center plane when viewed in the XZ plane.

Each of the drawing units MU1 to MU4 has a plurality of mirrors, a plurality of lenses, a rotating polygon mirror PM, a telecentric fθ lens system FT, etc., as disclosed in, for example, WO 2016/152758 and WO 2019/082850. The extensions of the center lines of the beams B1 and B3 emitted from the beam switching unit BD1B and incident on the corresponding odd-numbered drawing units MU1 and MU3, and the extensions of the center lines of the beams B2 and B4 emitted from the beam switching unit BD2B and incident on the corresponding even-numbered drawing units MU2 and MU4 are set to intersect with the rotation center line AXo of the rotating drum DR. Then, by rotation of the rotating polygon mirror PM of each of the drawing units MU1 to MU4, drawing lines SL1 to SL4 (SL3 and SL4 are not shown) as the trajectories of the spot light of each of the beams B1 to B4 scanned on the sheet substrate P are set parallel to the Y axis.

The patterns drawn by each of the drawing lines SL1 to SL4 are stitched together in the Y direction as the sheet substrate P moves in the longitudinal direction and exposed. Inside the drawing units MU1 to MU4, an orthogonal coordinate system XtYtZt is set, which is defined by the Zt axis set parallel to each of the incident beams B1 to B4, and the Xt and Yt axes which are orthogonal to the Zt axis. Therefore, the Yt axis of the orthogonal coordinate system XtYtZt is parallel to the Y axis of the orthogonal coordinate system XYZ, and the orthogonal coordinate system XtYtZt is inclined at a certain angle (θu) around the Y axis with respect to the XY plane of the orthogonal coordinate system XYZ.

Figure 2 is a perspective view showing a schematic internal configuration of the drawing unit MU1 as a representative of the drawing units MU1 to MU4 shown in Figure 1. The configuration of the drawing unit MU1 in Figure 2 is almost the same as that disclosed in, for example, International Publication No. 2016/152758, so it will be briefly described. The beam B1 from the beam switching unit BD1B includes at least one of the three beams B1a, B1b, and B1c made from the beams emitted from each of the three light source devices LS1A, LS1B, and LS1C when drawing a pattern. The three beams B1a, B1b, and B1c (collectively referred to as beam B1) are all parallel light beams with a diameter of 1 mm or less, and are incident on the mirror M10 in the drawing unit MU1 while maintaining a predetermined crossing angle with respect to each other. The beam B1 reflected by 90 degrees from the mirror M10 passes through a beam expander (expansion system) BEX consisting of lenses LGa and LGb arranged along the optical axis AXu1, and is then reflected by 90 degrees from the mirror M11 and enters the polarizing beam splitter PBS.

Beam B1 is linearly polarized in a direction perpendicular to the Zt axis direction, so it is efficiently reflected by the polarizing beam splitter PBS, reflected 90 degrees by mirror M12 and travels in the -Zt direction, and reflected 90 degrees by mirror M13 and travels in the +Xt direction. Beam B1 reflected by mirror M13 passes through a quarter-wave (λ/4) plate QP and a first cylindrical lens CYa, then is reflected by mirror M14 and reaches one reflecting surface Rp1 of the rotating polygon mirror PM. Beam B1 reflected by the reflecting surface Rp1 of the rotating polygon mirror PM is deflected in the XtYt plane by the rotation of the rotating polygon mirror PM, and enters a telecentric fθ lens system FT having an optical axis AXf1 parallel to the Xt axis.

Immediately after the fθ lens system FT, a mirror M15 is arranged to bend the optical axis AXf1 by 90 degrees, and the beam B1 emitted from the fθ lens system FT is reflected by the mirror M15 by 90 degrees so as to be parallel to the Zt axis. A second cylindrical lens CYb is arranged between the mirror M15 and the sheet substrate P, and the beam B1 (including at least one of the three beams B1a, B1b, and B1c) emitted from the fθ lens system FT is focused on the sheet substrate P as at least one of the spot lights SPa, SPb, and SPc. The spot lights SPa, SPb, and SPc are scanned one-dimensionally along a drawing line (scanning line) SL1 parallel to the Yt axis (Y axis) by the rotation of the rotating polygon mirror PM. 2 indicates the center line of rotation when the entire imaging unit MU1 including the optical members from the mirror M10 to the second cylindrical lens CYb is slightly rotated to tilt the imaging line SL1. An extension line of the line LE1 passes through the midpoint of the imaging line SL1 in the Yt direction and is set coaxially with the optical axis of the lens system in the beam switching unit BD1B that emits the beam B1 (the entire beams B1a, B1b, and B1c) incident on the mirror M10.

In the drawing unit MU1 shown in FIG. 2, the lens system LGc and the photoelectric sensor DT, which are arranged on the opposite side of the mirror M12 across the polarizing beam splitter PBS, receive the reflected light generated from the sheet substrate P by the projection of the spot light SP. By analyzing the waveform of the photoelectric signal from the photoelectric sensor DT, position information of the pattern already formed on the sheet substrate P can also be obtained. Also, in FIG. 2, the surface OPa inside the beam expander BEX is set to the rear focus of the lens LGa and the front focus of the lens LGb, and the beam B1 (at least one of B1a, B1b, and B1c) is focused at the position of the surface OPa into a beam waist with a cross-sectional dimension of several tens of μm. The surface OPa is ultimately in an optically conjugate relationship (imaging relationship) with the surface of the sheet substrate P. The cross-sectional dimensions (diameter, etc.) of each of the beams B1a, B1b, and B1c that pass through the lens LGb become parallel light beams expanded to several mm or more. The first cylindrical lens CYa and the second cylindrical lens CYb cooperate with the fθ lens system FT to correct positional fluctuations in the Xt direction of the spot light SP (the drawing line SL1) caused by differences in inclination between the reflecting surfaces of the rotating polygon mirror PM.

Figures 3A and 3B are exaggerated views showing the state of beam B1 (B1a, B1b, B1c) passing through beam expander BEX in drawing unit MU1 in Figure 2. In Figure 3A, since the configuration of beam expander BEX is the same in other drawing units MU2, MU3, and MU4, the beam entering beam expander BEX is generalized as Bn (n = 1 to 4), the three beams contained in beam Bn are generalized as Bna, Bnb, and Bnc (n = 1 to 4), and the optical axis is also generalized as AXun (n = 1 to 4). In this embodiment, the beams Bna (n = 1 to 4) incident on the lens LGa of the beam expander BEX from the light source devices LS1A and LS2A via the beam synthesis units BD1A and BD2A and the beam switching units BD1B and BD2B are set to be coaxial with the optical axis AXun (n = 1 to 4).

On the other hand, the beams Bnb (n = 1 to 4) generated from the light source devices LS1B and LS2B via the beam synthesis units BD1A and BD2A and the beam switching units BD1B and BD2B are inclined at a certain angle to the optical axis AXun (n = 1 to 4) in a plane including the optical axis AXun (n = 1 to 4) parallel to the XtYt plane and enter the lens LGa of the beam expander BEX. Similarly, the beams Bnc (n = 1 to 4) generated from the light source devices LS1C and LS2C via the beam synthesis units BD1A and BD2A and the beam switching units BD1B and BD2B are inclined at a certain angle to the optical axis AXun (n = 1 to 4) in a plane including the optical axis AXun (n = 1 to 4) parallel to the XtYt plane and enter the lens LGa of the beam expander BEX. Two beams Bnb, Bnc (n=1 to 4) incident on the beam expander BEX are incident on the lens LGa with symmetrical inclinations across the optical axis AXun (n=1 to 4) or the beam Bna (n=1 to 4).

Beam Bna (n = 1 to 4) incident on lens LGa is a parallel light beam, so on plane OPa, which corresponds to the pupil plane in beam expander BEX, a circular spot SPa' due to the beam waist of beam Bna is formed on optical axis AXun (n = 1 to 4) as shown in Figure 3B. Similarly, beams Bnb and Bnc (n = 1 to 4) incident on lens LGa are also parallel light beams, so on plane OPa in beam expander BEX, a spot SPb' due to the beam waist of beam Bnb and a spot SPc' due to the beam waist of beam Bnc are formed at symmetrical positions in the Yt direction across optical axis AXun (spot SPa') as shown in Figure 3B. As explained above, the cross-sectional shapes of the beams Bnb and Bnc are shaped into a slot shape (oblong ellipse) by the beam synthesis units BD1A and BD2A, respectively, so that the spots SPb' and SPc' are also each slot-shaped (oblong ellipse).

However, the long axis direction of the cross-sectional distribution of the slot-shaped (elliptical) spot SPb' is set to +45 degrees with respect to the Yt axis in the plane OPa, and the long axis direction of the cross-sectional distribution of the slot-shaped (elliptical) spot SPc' is set to -45 degrees with respect to the Yt axis in the plane OPa. Also, in the plane OPa, the center point of the cross-sectional distribution of the spot SPb' is set to be spaced apart from the center point (optical axis AXun) of the cross-sectional distribution of the circular spot SPa' by a distance Δyb', and the center point of the cross-sectional distribution of the spot SPc' is set to be spaced apart from the center point (optical axis AXun) of the cross-sectional distribution of the spot SPa' by a distance Δyc'.

The beams Bna, Bnb, and Bnc converged at the surface OPa enter the lens LGb of the beam expander BEX while diverging. At that time, in the optical path between the lens LGa and the lens LGb, the principal rays (central rays) of the beams Bna, Bnb, and Bnc are set to be approximately parallel to the optical axis AXun. The beam Bna that passes through the lens LGb becomes a circular parallel light beam with an expanded cross-sectional distribution and proceeds coaxially with the optical axis AXun. In addition, the beams Bnb and Bnc that pass through the lens LGb become a slot-shaped (elliptical) parallel light beam with an expanded cross-sectional distribution and proceed at an angle to the optical axis AXun. As explained in FIG. 2 above, the surface OPa is ultimately optically conjugate with the surface of the sheet substrate P. Therefore, the projected images of the spots SPa', SPb', and SPc' formed on the surface OPa are reduced and imaged onto the sheet substrate P as spot lights SPa, SPb, and SPc in Figure 2, respectively, via an imaging system consisting of lens LGb, first cylindrical lens CYa, fθ lens system FT, and second cylindrical lens CYb.

In this embodiment, as shown in FIG. 3B, the distance Δyb' between the spots SPa' and SPb' in the Yt direction, i.e., the center distance between the spot light SPa and the spot light SPb projected on the surface of the sheet substrate P in the Yt direction (main scanning direction), is proportional to the sine value of the inclination angle from the optical axis AXun of the principal ray (center ray) of the beam Bnb shown in FIG. 3A. Similarly, the distance Δyc' between the spots SPa' and SPc' on the surface OPa in the Yt direction, i.e., the center distance between the spot light SPa and the spot light SPc projected on the surface of the sheet substrate P in the Yt direction (main scanning direction), is proportional to the sine value of the inclination angle from the optical axis AXun of the principal ray (center ray) of the beam Bnc shown in FIG. 3A. Each of the spot lights SPa, SPb, and SPc projected onto the sheet substrate P has a cross-sectional shape and a positional relationship similar to those of the spots SPa', SPb', and SPc' in Fig. 3A. At least one of the spot lights SPa, SPb, and SPc is selected according to the shape of the pattern to be drawn and projected onto the sheet substrate P.

In the above configuration, a plane OPa conjugate with the surface of the sheet substrate P (the imaging plane of the spot lights SPa, SPb, SPc) is formed in the imaging unit MU1 (as are the other units MU2 to MU4), but the lens LGa of the beam expander BEX may be disposed outside the imaging unit MU1, and the mirror M10 in the imaging unit MU1 may be positioned between the plane OPa and the lens LGb. Furthermore, the beam expander BEX is not limited to a magnification system, and may be a 1x relay system.

Next, referring to Figures 4 and 5A to 5C, the configuration of the beam combining unit BD1A, which shapes the cross-sectional distribution of each of the three beams LB1a, LB1b, and LB1c from the light source device LS1 (LS1A, LS1B, and LS1C) shown in Figure 1 into a predetermined shape and combines them at a predetermined crossing angle, will be described. Figure 4 shows the schematic arrangement relationship between the light source devices LS1A, LS1B, and LS1C and the beam combining unit BD1A as viewed in the XY plane, and Figures 5A to 5C are diagrams that show the arrangement of optical members in the beam shape transformation unit 10B (or 10C) shown in Figure 4. The arrangement relationship and configuration of the light source devices LS2A, LS2B, and LS2C and the beam combining unit BD2A shown in Figure 1 are the same as those in Figure 4.

Light source devices LS1A, LS1B, and LS1C are fiber amplifier laser light sources of the same standard, and can oscillate a high-brightness ultraviolet pulse beam with an emission time of about several tens of picoseconds in response to each clock pulse of a common clock signal CLK. From the viewpoint of practicality and stability, the frequency of the clock signal CLK is set to about 400 MHz, for example. Each of light source devices LS1A, LS1B, and LS1C is supplied with drawing signals SDa, SDb, and SDc in the form of a bit stream, which serially reads out pixel bit data ("0" or "1") of drawing data (bitmap format) along a drawing line. The details of the drawing operation will be described later.

Beam LB1a from light source device LS1A (a parallel light beam with a circular cross section with a diameter of about 0.5 to 1 mm, corresponding to beam Bna supplied to drawing units MU1 and MU3) is incident on beam shape transformation unit 10A. As explained above in Figures 3A and 3B, the cross-sectional distribution of beam LB1a can remain almost circular, so beam shape transformation unit 10A may be omitted, but it is provided to match the optical path length of the lens system in beam shape transformation units 10B and 10C into which other beams LB1b and LB1c are incident. Beam shape transformation unit 10B receives beam LB1b from light source device LS1B (a parallel light beam with a circular cross section with a diameter of about 0.5 to 1 mm, corresponding to beam Bnb supplied to drawing units MU1 and MU3) and converts it into a parallel light beam with a slot-shaped (elliptical) cross-sectional distribution. Similarly, the beam shape deformation unit 10C receives beam LB1c (a parallel light beam with a circular cross-section having a diameter of approximately 0.5 to 1 mm, which corresponds to the beam Bnc supplied to the drawing units MU1 and MU3) from the light source device LS1C and converts it into a parallel light beam with a slot-shaped (oblong) cross-sectional distribution.

Beam LB1b traveling in the +X direction from beam shape transformation unit 10B is reflected at a right angle by mirror M2B in the XY plane and travels in the +Y direction, then reflected at a right angle by mirror M3B and travels in the +X direction. Beam LB1c traveling in the +X direction from beam shape transformation unit 10C is reflected at a right angle by mirror M2C in the XY plane and travels in the -Y direction, then reflected at a right angle by mirror M3C and travels in the +X direction. Beam LB1a from beam shape transformation unit 10A passes through the gap in the Y direction between mirrors M3B and M3C and travels in the +X direction in a parallel state with a constant Y-directional interval with other beams LB1b and LB1c. Three beams LB1a, LB1b, and LB1c (all parallel beams) arranged at a narrow interval in the Y direction in the XY plane are incident on quartz parallel plate 12A and wedge-shaped prisms 12B and 12C, respectively.

Beam LB1a passes through parallel plate 12A as it is positioned perpendicular to it, but the entrance and exit surfaces of wedge-shaped prisms 12B and 12C are not parallel to each other and form a certain angle (vertex angle) in the XY plane, so beam LB1b passing through prism 12B is refracted to approach beam LB1a in the XY plane, and beam LB1c passing through prism 12C is refracted to approach beam LB1a in the XY plane. After passing through parallel plate 12A and prisms 12B and 12C, the three beams LB1a, LB1b, and LB1c are reflected in the -Y direction by mirror M4, intersect with each other on plane OPm, and then move away from each other again before entering lens GK1. The front focus of lens GK1 is set to be at plane OPm.

Therefore, the beams LB1a, LB1b, and LB1c that pass through the lens GK1 converge to form beam waists at the rear focal point of the lens GK1, and the chief ray (center ray) of the beam LB1a becomes coaxial with the optical axis AXs of the lens GK1, and the chief ray (center ray) of the beam LB1b and the chief ray (center ray) of the beam LB1c become parallel to each other at a certain distance from the optical axis AXs. Note that, since the three beams LB1a, LB1b, and LB1c (each a parallel beam) overlap at the optical axis AXs in the plane OPm, the plane OPm is a superposition of the intensity distribution of the beam LB1a with a circular cross section and the intensity distribution of the beams LB1b and LB1c with a slot-shaped (elliptical) cross section with a major axis direction of approximately 90 degrees. In addition, in this embodiment, the vertex angles of the prisms 12B and 12C set the inclination angles of the two beams Bnb and Bnc incident on the lens LGa of the beam expander BEX of the drawing unit MU1 (MU2 to MU4) shown in Figures 3A and 3B with respect to the optical axis AXun.

4, the clock signal CLK is output from a clock generating unit 100A provided in the control device 100, and each of the drawing signals SDa, SDb, and SDc is output from a drawing data storage unit 100B provided in the control device 100. In the control device 100, there are provided a switching control unit 100C that applies a drive signal to an acousto-optical modulation element provided in the beam switching unit BD1B (BD2B is also the same) shown in FIG. 1, a drawing unit control unit 100D that controls the rotary motors of each polygon mirror PM in the drawing units MU1 and MU3 (MU2 and MU4 are also the same) shown in FIG. 2 and receives an origin signal (timing signal) generated for each reflecting surface of the polygon mirror PM, and a rotating drum control unit 100E that controls the rotary motor of the rotating drum DR shown in FIG. 1 and receives measurement information from an encoder that measures its rotation angle position.

4 starts reading out the drawing data string (serial bit string) for one scan along the drawing lines SL1, SL3 (SL2, SL4 as well) by the spot lights SPa, SPb, SPc in response to an origin signal (timing signal) received by the drawing unit control unit 100D, and performs address switching of the drawing data string for one scan to be read out based on the measurement information of the encoder received by the rotating drum control unit 100E. The switching control unit 100C also controls the on/off modulation of the acousto-optical modulation element in the beam switching unit BD1B (BD2B) in response to the origin signal (timing signal) from each of the drawing units MU1, MU3 (MU2, MU4) received by the drawing unit control unit 100D.

Figures 5A to 5C show the detailed optical configuration of each of the beam shape transformation units 10B and 10C, and each basically comprises a beam expansion system OM1, a beam compression system OM2, and a beam reduction system OM3. The beam expansion system OM1 and beam reduction system OM3 on the beam shape transformation unit 10B side, which are arranged along the optical axis AXb, and the beam expansion system OM1 and beam reduction system OM3 on the beam shape transformation unit 10C side, which are arranged along the optical axis AXc, are all similarly configured using the same optical members. Also, Figures 5A and 5B are views of the beam shape transformation units 10B and 10C as viewed from a direction rotated 90 degrees around the optical axis AXb (AXc).

The beam expansion system OM1, into which the beam LB1b (LB1c) from the light source device LS1B (LS1C) is incident, is composed of a spherical negative lens 10G1 and a spherical positive lens 10G2, and converts the cross-sectional distribution diameter of the beam LB1b (LB1c) into a parallel beam expanded to a distribution BVa several times larger. The beam compression system OM2 is composed of two cylindrical lenses 10G3 and 10G4 arranged along the optical axis AXb (AXc). The cylindrical lens 10G3 is a convex lens that has no refractive power in the direction of the generating line Ds and has positive refractive power in the direction perpendicular to each of the generating line Ds and the optical axis AXb (AXc). Moreover, the cylindrical lens 10G4 is a concave lens that has no refractive power in the direction of the generating line Ds and has negative refractive power in the directions perpendicular to both the generating line Ds and the optical axis AXb (AXc).

The expanded beam LB1b (LB1c) from the beam expansion system OM1 passes through the cylindrical lenses 10G3 and 10G4 as a parallel beam, but as shown in FIG. 5A, in a plane parallel to the generating line Ds including the optical axis AXb (AXc), the cylindrical lenses 10G3 and 10G4 simply function as parallel plates, so that the beam enters the next beam reduction system OM3 in the parallel state. On the other hand, in a plane including the optical axis AXb (AXc) and perpendicular to the generating line Ds, the expanded beam LB1b (LB1c) enters the next beam reduction system OM3 in a parallel state in which the width of the beam is reduced by the refractive power of the cylindrical lenses 10G3 and 10G4. Therefore, the distribution in the cross section of the beam LB1b (LB1c) emerging from the cylindrical lens 10G4 is shaped like a slot-shaped (oblong) distribution BVb.

The beam reduction system OM3 is composed of a spherical positive lens 10G5 and a spherical negative lens 10G6 arranged along the optical axis AXb (AXc), and converts the cross-sectional distribution of the beam LB1b (LB1c) emitted as a parallel beam from the beam compression system OM2 into an isotropically reduced parallel beam. The cross-sectional distribution of the beam LB1b (LB1c) emitted from the negative lens 10G6 is a slot-shaped (elliptical) shape with the direction of the generatrix Ds as the major axis, and as an example, the width in the major axis direction is set to be approximately the same as the diameter of the circular cross-sectional distribution of the beam LB1a emitted from the beam shape transformation unit 10A. The width in the minor axis direction of the cross-sectional distribution of the beam LB1b (LB1c) emitted from the negative lens 10G6 is set to approximately 1/4 to 1/6 of the width in the major axis direction.

In the configurations of Figures 5A to 5C, the direction of the generatrix Ds of the cylindrical lenses 10G3 and 10G4 in the beam compression system OM2 on the beam shape transformation unit 10B side and the direction of the generatrix Ds of the cylindrical lenses 10G3 and 10G4 in the beam compression system OM2 on the beam shape transformation unit 10C side are set to form approximately 90 degrees when viewed in a plane perpendicular to the optical axes AXb and AXc. As shown in Figure 5C, for example, when viewed from the mirrors M2B and M2C in Figure 4, the generatrix Ds of the cylindrical lenses 10G3 and 10G4 on the beam shape transformation unit 10B side is rotated 45 degrees counterclockwise from the Y axis, and the generatrix Ds of the cylindrical lenses 10G3 and 10G4 on the beam shape transformation unit 10C side is rotated 45 degrees clockwise from the Y axis. As a result, the major axis direction of the slot-shaped (oblong) cross-sectional distribution of beam LB1b emitted from beam shape deformation section 10B and the major axis direction of the slot-shaped (oblong) cross-sectional distribution of beam LB1c emitted from beam shape deformation section 10C form an angle of approximately 90 degrees.

In addition, the beam shape transformation unit 10A shown in FIG. 4 is configured by inserting a simple parallel plate (made of quartz) instead of the beam compression system OM2 (two cylindrical lenses 10G3, 10G4) shown in FIG. 5A and FIG. 5B in order to match the optical path length with the other beam shape transformation units 10B and 10C. Therefore, the beam shape transformation unit 10A is composed of a beam expansion system OM1, a parallel plate, and a beam reduction system OM3. However, if it is not necessary to match the optical path length, the beam shape transformation unit 10A itself can be omitted. Note that the entrance surface or exit surface of either one or both of the two cylindrical lenses 10G3 and 10G4 may not be a perfect cylindrical surface like an aspherical lens, but may be an approximate cylindrical surface approximated by a higher-order function, etc.

In this embodiment, as shown in FIG. 5C, the direction of the generatrix Ds of the two cylindrical lenses 10G3 and 10G4 is set at about 45 degrees from the Y axis, but by holding the two cylindrical lenses 10G3 and 10G4 integrally in a lens barrel and configuring the lens barrel to be rotatable around the optical axis AXb (AXc), the major axis direction of the slot-shaped (elliptical) spot lights SPb and SPc can be set (rotated) in any direction on the sheet substrate P. Note that the configuration of the beam combining unit BD2A shown in FIG. 1 is an arrangement in which the beam combining unit BD1A shown in FIG. 4 is rotated 180 degrees around an axis parallel to the Z axis, and the beam shape transformation units 10A, 10B, and 10C in the beam combining unit BD2A are also configured in the same manner as the optical members shown in FIGS. 5A to 5C.

In the configuration of Figures 5A to 5C, a beam expansion system OM1 is provided before the beam compression system OM2 and a beam reduction system OM3 is provided after it, but these may be omitted and the beam shape deformation units 10B, 10C may be constructed using only the beam compression system OM2. However, if it is necessary to adjust the dimensions (particularly the length in the major axis direction) of the slot-shaped (or oblong) spot lights SPb, SPc that are ultimately projected onto the sheet substrate P, the size of the spot lights SPb, SPc can be set to the desired dimensions by providing a beam expansion system OM1 and a beam reduction system OM3.

Next, the detailed configuration of the beam switching units BD1B and BD2B shown in FIG. 1 will be described with reference to FIG. 6. The beam switching units BD1B and BD2B have the same basic configuration, and in FIG. 1, the beam switching unit BD2B is arranged by rotating the entire beam switching unit BD1B by 180 degrees around an axis parallel to the Z axis. Therefore, the configuration of the beam switching unit BD1B will be described with reference to FIG. 6 as a representative. FIG. 6 is a perspective view showing a schematic optical path of the beam switching unit BD1B from the lens GK1 in the beam synthesis unit BD1A shown in FIG. 4 to the odd-numbered drawing units MU1 and MU3, and the orthogonal coordinate system XYZ is set to the same as the coordinate system XYZ in FIG. 1.

Beam LB1 (LB1a, LB1b, LB1c) from lens GK1 travels in the -Y direction parallel to the optical axis AXs, is reflected perpendicularly to the -Z direction by beam splitter M40, is deflected by mirrors M41 and M42 to travel in the -X direction, and enters lens GK2. Beam splitter M40 transmits a light component of beam LB1 (LB1a, LB1b, LB1c) that is a few percent or less of the light amount, and makes it into measurement beam MLB, and reflects the remaining light component in the -Z direction. Measurement beam MLB is entered into a beam monitor system (not shown), which measures the light amount (light intensity or light energy) of each of the three beams LB1a, LB1b, LB1c, as well as fluctuations in the positional relationship between the beams.

In FIG. 6, the plane OPs between the beam splitter M40 and the mirror M41 is the position of the rear focal point of the lens GK1, and on the plane OPs, the beam waists (spots) of the beams LB1a, LB1b, and LB1c are aligned in the Y direction as shown in FIG. 3B. The chief ray (central ray) of the beam LB1a, which travels as a diverging beam from the plane OPs, is coaxial with the optical axis AXs, and the chief ray (central ray) of each of the beams LB1b and LB1c, which travel as a diverging beam from the plane OPs, is both parallel to the optical axis AXs. The beams LB1a, LB1b, and LB1c that pass through the lens GK2, which is arranged so that the position of the front focal point is the plane OPs, are each converted into a parallel beam and tilted so as to intersect with each other at a predetermined angle in the XY plane in FIG. 6.

The beams LB1a, LB1b, and LB1c that pass through the lens GK2 in the -X direction are incident on the first-stage acousto-optical modulation element AM3. The acousto-optical modulation element AM3 is disposed at the rear focal point of the lens GK2 so that the three beams LB1a, LB1b, and LB1c intersect in a plane parallel to the XY plane in the crystal of the acousto-optical modulation element AM3. Therefore, the plane OPm shown in FIG. 4 and the first-stage acousto-optical modulation element AM3 are in a conjugate relationship due to the relay optical system formed by the lenses GK1 and GK2. The acousto-optical modulation element AM3 is disposed so that the conditions for Bragg diffraction are satisfied for the incident beam, and the diffraction direction is in the -Z direction. Then, while the acousto-optical modulation element AM3 is in an on state (a state in which a high-frequency drive signal is being applied), a zero-order beam (parallel light beam) and a first-order diffracted beam (parallel light beam) of each of the incident beams LB1a, LB1b, and LB1c are generated from the acousto-optical modulation element AM3. The zero-order beam and the first-order diffracted beam are reflected back by mirrors M43 and M44, and travel in the +X direction to be incident on the lens GK3.

The position of the front focal point of the lens GK3 is set within the crystal of the acousto-optical modulator AM3, so that the chief ray (center ray) of each of the zeroth order beams of the beams LB1a, LB1b, and LB1c traveling in the +X direction from the lens GK3 and the chief ray (center ray) of each of the first order diffracted beams are parallel to the optical axis AXs and are separated from each other in the YZ plane (plane perpendicular to the optical axis AXs). Each of the first order diffracted beams of the beams LB1a, LB1b, and LB1c is selectively reflected in the -Z direction by the 45-degree reflecting surface of the epi-mirror IM3 provided at the position of the rear focal point of the lens GK3, and each of the zeroth order beams of the beams LB1a, LB1b, and LB1c passes through the space above the epi-mirror IM3 in the +Z direction. When the acousto-optical modulation element AM3 is used under Bragg diffraction conditions, the amount of light of the generated first-order diffracted beam is 80 to 90% of the incident beam, and the remainder is the amount of light of the zeroth-order beam.

Now, referring to Figures 7A and 7B, the state of each beam in the optical path from the acousto-optical modulation element AM3 to the incident light mirror IM3 will be described in detail. Figure 7A shows the optical path as viewed in the XY plane, and Figure 7B shows the optical path as viewed in the XZ plane. As shown in Figure 7A, when viewed in the XY plane, each of the beams LB1a, LB1b, and LB1c, which are parallel light beams, intersect at position Pe (the rear focus of lens GK2 and the front focus of lens GK3) within the crystal of the acousto-optical modulation element AM3. As shown in FIG. 7B, each of the beams LB1a, LB1b, and LB1c incident on the acousto-optical modulation element AM3 is aligned along the optical axis AXs when viewed in the XZ plane, but when the acousto-optical modulation element AM3 is in the on state, a zeroth-order beam B3ao and a first-order diffracted beam B3a of the beam LB1a, a zeroth-order beam B3bo and a first-order diffracted beam B3b of the beam LB1b, and a zeroth-order beam B3co and a first-order diffracted beam B3c of the beam LB1c are generated, and the first-order diffracted beam B3a, the first-order diffracted beam B3b, and the first-order diffracted beam B3c are deflected in the -Z direction at a predetermined diffraction angle with respect to their respective zeroth-order beams.

When viewed in the XY plane, the zeroth-order beam B3ao and the first-order diffracted beam B3a, the zeroth-order beam B3bo and the first-order diffracted beam B3b, and the zeroth-order beam B3co and the first-order diffracted beam B3c are each vertically overlapped. After passing through the lens GK3, the zeroth-order beams B3ao, B3bo, and B3co each become a convergent beam and travel parallel to the optical axis AXs in the XY plane, and then become a beam waist (spot) at the plane Pso (the rear focal position of the lens GK3) where the reflecting surface of the epi-mirror IM3 is located, and then travel as a divergent beam in the space above the epi-mirror IM3 in the +Z direction. Each of the first-order diffracted beams B3a, B3b, and B3c that pass through lens GK3 also becomes a convergent light beam, travels parallel to the optical axis AXs along an optical path that is a certain distance away from the optical axis AXs in the -Z direction, becomes a beam waist at plane Pso, and is reflected in the -Z direction by the reflecting surface of the epi-illumination mirror IM3.

The first-order diffracted beams B3a, B3b, and B3c (whose central rays are parallel to each other) reflected by the reflecting surface of the epi-mirror IM3 become divergent light beams and travel toward the imaging unit MU3. In Figures 7A and 7B, the optical axis AXu3 of the optical path from the epi-mirror IM3 toward the imaging unit MU3 corresponds to the optical axis AXun of the beam expander BEX (lenses LGa, LGb) described in Figures 2, 3A, and 3B.

Returning to the explanation of FIG. 6, when the acousto-optical modulation element AM3 is in the off state (high-frequency drive signal is not applied), each of the three beams LB1a, LB1b, and LB1c incident on the acousto-optical modulation element AM3 is not diffracted but passes through the lens GK3 and reaches the mirror M45 after passing through the space above the incident light mirror IM3 along the same optical path as the zero-order beams B3ao, B3bo, and B3co shown in FIG. 7A and FIG. 7B. The mirror M45 reflects the three beams LB1a, LB1b, and LB1c (whose central rays are parallel to each other in the XY plane) in the -Y direction toward the mirror M46. The mirror M46 further reflects the beams LB1a, LB1b, and LB1c in the -X direction toward the lens GK4. The position of the front focal point of the lens GK4 is set to the position of the beam waist formed on the reflecting surface of the epi-illumination mirror IM3 or in its immediate vicinity (the same position as the plane Pso in FIGS. 7A and 7B).

Beam LB1a that passes through lens GK4 is converted into a parallel beam and travels coaxially with optical axis AXs, while beams LB1b and LB1c that pass through lens GK4 are each converted into a parallel beam and travel at an angle in the XY plane so as to intersect with beam LB1a (optical axis AXs). Acousto-optical modulation element AM1 is disposed at the rear focal position of lens GK4, and the three beams LB1a, LB1b, and LB1c (parallel beams) emitted from lens GK4 intersect within the crystal of acousto-optical modulation element AM1, similar to the state shown in Figure 7A above. When the acousto-optical modulation element AM1 is in the on state, the zeroth-order beam B1ao and first-order diffracted beam B1a of beam LB1a, the zeroth-order beam B1bo and first-order diffracted beam B1b of beam LB1b, and the zeroth-order beam B1co and first-order diffracted beam B1c of beam LB1c are emitted from the acousto-optical modulation element AM1 in a state similar to that shown in Figures 7A and 7B.

The zero-order beams B1ao, B1bo, and B1co (all parallel beams) emitted from the acousto-optical modulation element AM1 and the first-order diffracted beams B1a, B1b, and B1c (all parallel beams) deflected in the -Z direction at a predetermined diffraction angle are folded back in the XY plane by mirrors M47 and M48, proceed in the +X direction, and enter the lens GK5. The position of the front focal point of the lens GK5 is set within the crystal of the acousto-optical modulation element AM1, and the epi-mirror IM1 similar to the epi-mirror IM3 is placed at the position of the rear focal point of the lens GK5. As in the state described in Figures 7A and 7B above, when the acousto-optical modulation element AM1 is in the on state, each of the first-order diffracted beams B1a, B1b, and B1c is converged to a beam waist on the 45-degree reflecting surface of the epi-mirror IM1 or a position very close to it, and is reflected in the -Z direction along the optical axis AXu1 on the drawing unit MU1 side. In the above configuration, the acousto-optic modulation element AM1 and the acousto-optic modulation element AM3 are set in a conjugate relationship with each other by a unit-magnification relay optical system (imaging system) consisting of two lenses GK3 and GK4.

The beams Bna, Bnb, Bnc (n = 1 to 4) reflected by the odd-numbered incident mirrors IM1, IM3 (even-numbered incident mirrors IM2, IM4) shown in Figures 6, 7A, and 7B each have a central ray parallel to the optical axis AXun (n = 1 to 4), but are divergent beams. Therefore, in order to convert each of the beams Bna, Bnb, Bnc into parallel beams that intersect with each other, a lens GK6 as shown in Figure 8 is provided at the final stage of the optical path of the beam switching units BD1B, BD2B. Figure 8 is a perspective view showing the state of the beams Bna, Bnb, Bnc incident on the lens LGa of the beam expander BEX of each of the drawing units MU1 to MU4 from each of the beam switching units BD1B, BD2B.

In Figure 8, beam Bna (divergent light beam) incident on lens GK6 coaxially with optical axis AXun becomes a parallel light beam (diameter about 1 mm) from lens GK6, is reflected at a right angle in the -Xt direction by mirror M10 shown in Figure 2, and is incident on lens LGa of beam expander BEX coaxially with optical axis AXun. Plane Pe' set in the optical path between lenses GK6 and LGa is the rear focal position of lens GK6 and the front focal position of lens LGa.

The front focus of the lens GK6 is set at the position of the plane Pso shown in Figures 7A and 7B. Therefore, each of the beams Bnb and Bnc (divergent light beams) incident on the lens GK6 is converted from the lens GK6 into a parallel light beam (diameter about 1 mm), intersects with the position of the optical axis AXun in the plane Pe', is reflected by the mirror M10, passes through the lens LGa of the beam expander BEX, and is guided into the drawing unit MUn (n = 1 to 4). As explained in Figures 3A and 3B above, each of the beams Bna, Bnb, and Bnc that pass through the lens LGa of the beam expander BEX advances parallel to the optical axis AXun and converges to become spots SPa', SPb', and SPc' on the plane OPa, respectively.

As an example of the operation of drawing a pattern on a sheet substrate P by the pattern exposure apparatus EX configured as shown in Figures 1 to 8 above, the case of drawing a pattern as shown in Figures 9A and 9B will be described. Figure 9A shows line and space (L&S) patterns PT1, PT2, PT3 consisting of eight lines exposed onto the sheet substrate P, and Figure 9B shows pixel map (bitmap) information on the drawing data of an enlarged partial area Acc of the pattern. Such pixel map information is pre-stored in the drawing data storage unit 100B in the control device 100 shown in Figure 4 above.

In FIG. 9A, pattern PT1 is an L&S pattern in which eight lines (black lines) with a line width of 20 μm extending linearly in the Xt direction (sub-scanning direction) are arranged with a space width of 20 μm in the Yt direction (main scanning direction), and pattern PT3 is an L&S pattern in which eight lines (black lines) with a line width of 10 μm extending linearly in the Yt direction are arranged with a space width of 10 μm in the Xt direction. Pattern PT2 is an L&S pattern in which the eight lines of patterns PT1 and PT3 are connected by eight lines inclined at about 45 degrees to the Xt direction and the Yt direction. Each of the eight lines (black lines) of pattern PT2 has a line width of about 10 μm, and the space width is set to about 21.3 μm. Such configurations of patterns PT1, PT2, and PT3 are often provided as wiring layers on electronic devices.

9A, the diagonal line PT2a of the pattern PT2 and the straight line PT3a of the pattern PT3 extending in the Yt direction are connected at about 135 degrees, and the diagonal line PT2b of the pattern PT2 and the straight line PT3b of the pattern PT3 extending in the Yt direction are connected at about 135 degrees. In this case, as shown in FIG. 9B, on the pixel map (bitmap) corresponding to the region Acc, the size of one pixel Pic is defined as, for example, a square of 2×2 μm on the sheet substrate P, so that the line width direction (Xt direction) of each of the straight lines PT3a and PT3b is set to 5 pixels (5 Pic) as shown by hatching. And the space width in the Xt direction between the straight lines PT3a and PT3b is also set to 5 pixels (5 Pic). On the other hand, the 45-degree diagonal lines PT2a and PT2b are set to a line width of approximately 10 μm, so that the dimension in the Yt direction (or Xt direction) is approximately 14.1 μm, and seven pixels (7 Pics) indicated by hatching are set in each of the Yt and Xt directions.

In FIG. 9B, pixel Pic is defined as one bit of "0" or "1" in the drawing data. For example, when pixel Pic is "1", one of spot lights SPa, SPb, SPc is pulse-exposed on sheet substrate P in response to clock pulses of clock signal CLK. In this case, when the main scanning direction of the spot light is set to the -Yt direction (from left to right in FIG. 9B), the drawing data in area Acc in FIG. 9B is a data string for one drawing line of drawing data, for example, pixel bit information in data strings AL1 and AL2 in FIG. 9B, is read out sequentially from left to right in response to clock signal CLK. The read bit information is applied to light source devices LS1A, LS1B, and LS1C as bit stream-like drawing signals SDa, SDb, and SDc shown in FIG. 4. Each of the data strings AL1 and AL2 has a bit string of pixels equal to Lmy/Ypi, where Lmy (μm) is the maximum length in the Yt direction of the drawing line SLn (n=1 to 4) shown in Figures 1 and 2, and Ypi (μm) is the pixel dimension of the pixel Pic in the Yt direction.

In this embodiment, for example, when any of the odd-numbered drawing units MU1, MU3, MU5 exposes a pattern including diagonal lines as shown in Figure 9A, the pattern is drawn by selectively switching at high speed between a circular spot light SPa formed by a beam LB1a from light source device LS1A, a slot-shaped spot light SPb formed by a beam LB1b from light source device LS1B, and a slot-shaped spot light SPc formed by a beam LB1c from light source device LS1C. Therefore, in this embodiment, when the drawing data as shown in FIG. 9B is taken as the basic drawing data (basic data string ALx) and the address of the pixels arranged in the Xt direction is taken as x, three data strings are generated and stored: a first data string ALxa corresponding to the pattern portion to be drawn by the circular spot light SPa, a second data string ALxb corresponding to the pattern portion to be drawn by the slot-shaped spot light SPb inclined at -45 degrees, and a third data string ALxc corresponding to the pattern portion to be drawn by the slot-shaped spot light SPc inclined at +45 degrees.

Fig. 10 is a diagram for explaining an operation in the case of drawing a part of one oblique line pattern in the pattern PT2 shown in Fig. 9B as an example. In Fig. 10, the dimension Xpi in the Xt direction and the dimension Ypi in the Yt direction of the pixel Pic on the sheet substrate P are set to Xpi = Ypi. The effective dimension (diameter) of the circular spot light SPa is set to a value equal to or slightly larger than the dimensions Xpi, Ypi of the pixel Pic. The effective diameter means a diameter at a level of 1/ ^e2 or 1/2 of the peak intensity when the intensity distribution of the spot light SPa is a Gaussian distribution or an approximate Gaussian distribution. In addition, the effective dimensions in the major axis direction of each of the slot-shaped (elliptical) spot lights SPb, SPc inclined at 45 degrees are set to be equal to or slightly larger than the dimensions Xpi, Ypi of the square pixel Pic, or the diagonal dimensions of the pixel Pic (approximately 1.4 times Xpi, Ypi).

Furthermore, the center of the spot light SPb is set to be spaced apart from the center of the spot light SPa by a distance ΔYb in the -Yt direction, and the center of the spot light SPc is set to be spaced apart from the center of the spot light SPa by a distance ΔYc in the +Yt direction. In FIG. 10, for ease of explanation, the distances ΔYb and ΔYc are set to be equal, 2·Ypi, which is the distance between two pixels, but the distances ΔYb and ΔYc may be more than two pixels if they are known in advance. The effective dimensions (diameter) of the circular spot light SPa may be within a range of ±50% (preferably within a range of ±30%) of the dimensions of the pixel Pic set on the sheet substrate P.

As shown in FIG. 10, for a diagonal line pattern tilted at -45 degrees, a slot-shaped spot light SPb tilted at -45 degrees is selected, and the spot light SPb is pulsed in response to each clock pulse of the clock signal CLK (400 MHz). The pulsed irradiation of the spot light SPb (as well as the other spot lights SPa, SPc) is set so that one pixel Pic receives two pulses in the main scanning direction (Yt direction). Specifically, the rotation speed of the polygon mirror PM is set so that the spot light SPb (SPa, SPc) moves by 1/2 the dimension Ypi of the pixel Pic in the Yt direction during the period Tck (2.5 nS) of the clock signal CLK, and the scanning speed of the spot light SPb (SPa, SPc) is set to 0.5·Ypi/Tck (μm/nS).

Similarly, for the Xt direction (sub-scanning direction), the movement speed of the sheet substrate P in the Xt direction, i.e., the rotation speed of the rotating drum DR (see FIG. 1), is set so that two drawing lines SL1a and SL1b are set for one pixel Pic, as shown in FIG. 10. The drawing lines SL1a and SL1b are the result of the beam B1b (as well as the other beams B1a and B1c) incident on the drawing unit MU1 (as well as the other drawing units MU2 to MU4) being reflected by each of the adjacent reflecting surfaces in the rotation direction of the polygon mirror PM. Therefore, when the polygon mirror PM has eight reflecting surfaces, the speed relationship is set so that the sheet substrate P moves 1/2 the dimension Xpi of the pixel Pic in the Xt direction while the polygon mirror PM rotates 45°.

In Fig. 10, the pulse emission of light source device LS1B is controlled according to the bit value "1" in the drawing signal SDb based on pixel bit information and the clock pulse of the clock signal CLK so that a slot-shaped spot light SPb tilted at -45 degrees is irradiated to the diagonal line pattern portion (pixels Pic marked with black dots) along the data string AL1. Meanwhile, as shown in Fig. 10, the pixel bit information of the drawing signals SDa and SDc applied to each of the other light source devices LS1A and LS1C has a bit value of "0", so that pulse irradiation by the spot lights SPa and SPc is not performed.

As shown in Figure 10, by selectively setting pixel bit information for the same pixel contained in each of the drawing signals SDa, SDb, SDc to either bit value "0" or "1", one of the three spot lights SPa, SPb, SPc can be selected to reduce jagged edges of the exposed pattern for each of the patterns PT2 including the diagonal lines shown in Figure 9A, and the patterns PT1 and PT3 including straight lines in the Xt or Yt direction.

Figure 11 is a diagram explaining the state of pixel bit information (drawing signals SDa, SDb, SDc) corresponding to each of pixel data strings AL1 and AL2 in area Acc in Figure 9B, among the drawing data for drawing a portion of the diagonal line shown in Figure 9B. Data string AL1 or AL2 includes both diagonal line portions PT2a, PT2b and straight line portion PT3b in a pixel row aligned in the main scanning direction. For both data strings AL1 and AL2, area Acc in Figure 9B is defined by 37 pixels in the main scanning direction. Assuming that the drawing signal SDa for drawing a pattern with the spot light SPa is generated from the designed data string AL1, the first pixel to the fourth pixel at the left end in FIG. 9B, which is the first pixel to be read out, stores a bit value of "0" (non-drawing), the fifth pixel to the eleventh pixel (the seven hatched pixels) stores a bit value of "1" (drawing) corresponding to the diagonal line PT2a, the twelfth pixel to the 27th pixel stores a bit value of "0" (non-drawing), and the 28th pixel to the 37th pixel store a bit value of "1" (drawing) corresponding to the edge pixel to the left of the diagonal line PT2b and the straight line PT3b.

Similarly, in the design data column AL2, which is shifted one step in the sub-scanning direction from the design data column AL1, a bit value of "1" (drawn) corresponding to the diagonal line PT2a is stored in the 6th to 12th pixels (hatched pixels Pic), a bit value of "0" (not drawn) is stored in the 13th to 28th pixels, and a bit value of "1" (drawn) corresponding to the edge pixel to the left of the diagonal line PT2b and the straight line PT3b is stored in the 29th to 37th pixels.

As explained above in Figure 10, the diagonal line PT2a inclined at -45 degrees is exposed with the spot light SPb, so the bit value "1" is set for the 3rd to 9th pixels (7 pixels) in the data string on the drawing signal SDb corresponding to the designed data string AL1 (drawing signal SDa). As explained in Figure 10, the spot light SPb is set at a position where it draws a pattern two pixels (ΔYb) ahead of the spot light SPa in the main scanning direction, so the data string generating the drawing signal SDb has bit values set so that it is generally two pixels (2 bits) ahead of the designed data string AL1 generating the drawing signal SDa.

Furthermore, in the design data string AL1, a bit value of "1" is stored at the 28th pixel for the edge pixel on the left side of the diagonal line PT2b, and bit values of "1" corresponding to the straight line PT3b are stored at the subsequent 29th pixel and onwards. In order to expose the 28th pixel in the design data string AL1 with two pulses of the spot light SPb, in the drawing signal SDa generated by the design data string AL1, the 28th pixel is set to a bit value of "0" (non-drawing), and the subsequent 29th pixel and onwards are set to a bit value of "1". Additionally, in the data string generating the drawing signal SDb, a bit value of "1" is set at the 26th pixel, which is two pixels preceding the 28th pixel in the design data string AL1.

In addition, in the area Acc shown in FIG. 9B, since there is no pattern of a diagonal line inclined at +45 degrees, all pixels in the data string generating the drawing signal SDc in FIG. 11 are set to a bit value of "0" (non-drawing). When drawing a diagonal line (or a diagonal edge portion) inclined at +45 degrees with the spot light SPc, a bit value of "1" is set to a specific pixel in the data string generating the drawing signal SDc. In this case, since the spot light SPc is set to a position where it draws a pattern by delaying it by two pixels (ΔYc) in the main scanning direction from the spot light SPa, the bit value of the data string generating the drawing signal SDc is set so that it is behind (delayed) by two pixels (two bits) overall with respect to the designed data string AL1 generating the drawing signal SDa.

Similarly, for data string AL2, the bit value "1" is set for the 4th to 10th pixels (7 pixels) in the data string on the drawing signal SDb, which precedes the designed data string AL2 (drawing signal SDa) by two pixels. Furthermore, since the edge pixel on the left side of the diagonal line PT2b is located at the 29th pixel in the designed data string AL1, the bit value "1" is set for the 27th pixel, which precedes by two pixels, in the data string on the drawing signal SDb, and the bit value "0" is set for the 28th pixel and thereafter. Meanwhile, in the drawing signal SDa generated by the designed data string AL1, the bit value "0" is set for the 29th pixel, and the bit value "1" is set for the subsequent 30th pixel and thereafter.

As described above, each of the drawing signals SDa, SDb, and SDc is generated by bit-shifting the data string (AL1, AL2, etc.) read out every two clock pulses of the clock signal CLK during one scan of the spot light SPa, SPb, and SPc by an amount corresponding to the relative interval ΔYb, ΔYc in the main scanning direction of the spot light SPa, SPb, and SPc. Since the spot light SPb and the spot light SPc are shifted by an interval (ΔYb + ΔYc) in the main scanning direction, the data string generating the drawing signal SDb and the data string generating the drawing signal SDc are shifted by the number of bits (here, 4 pixels) corresponding to the interval (ΔYb + ΔYc).

As described above, in this embodiment, even if the pattern to be drawn includes a diagonal line pattern or a diagonal edge portion, drawing with a circular spot light SPa and drawing with a tilted slot-shaped (long elliptical) spot light SPb or SPc can be accurately switched in pixel units on the drawing data. In particular, by roughly aligning the effective diameter φs of the spot light SPa and the dimensions (length in the major axis direction) of the spot lights SPb and SPc without significantly changing them, it is possible to reduce the jaggedness that occurs in the diagonal line pattern and the diagonal edge portion, and to accurately maintain the line width. In addition, in the configuration of the beam synthesis unit BD1A (BD1B) shown in Figures 4 and 5A to 5C, the polarization states (directions of linear polarization) of the beams LB1a, LB1b, and LB1c from each of the three light source devices LS1A, LS1B, and LS1C (LS2A, LS2B, and LS2C) can be aligned and made to enter the first-stage acousto-optical modulation element AM3 (AM4).

In addition, in this embodiment, the major axis direction of each of the slot-shaped (elliptical) spot lights SPb and SPc is inclined at 45 degrees with respect to the main scanning direction (or sub-scanning direction), because wiring and pattern edges inclined at 45 degrees are often used in pattern designs for many electronic devices, especially wiring designs. However, even if the pattern to be drawn includes wiring (line patterns) or pattern edge portions with an angle β (tilt angle with respect to the main scanning direction or sub-scanning direction) that is different from 45 degrees, if the angle β is in the range of |β-45°|≦20°, that is, in the range of 25°≦β≦65°, the selective exposure with the slot-shaped (elliptical) spot lights SPb and SPc inclined at 45 degrees can reduce the jaggedness of the diagonal edge portion.

Furthermore, in this embodiment, as described in Figures 5A to 5C above, the major axis direction of each of the spot lights SPb and SPc can be set to any direction (in practice, a range of 0° to 90° is sufficient) by rotating around the optical axis AXb (AXc) of the beam compression system OM2. Therefore, the angles of each of the diagonal lines and diagonal edge portions that appear in the pattern for the electronic device exposed on the sheet substrate P relative to the main scanning direction (Yt direction) can be tallied, the most frequent angle (high frequency angle) can be found, and the major axis direction of each of the spot lights SPb and SPc can be set to correspond to that high frequency angle. Note that instead of rotating the beam compression system OM2, a trapezoidal prism or an image rotator with three reflecting surfaces can be provided after the beam compression system OM2 to rotate it around the optical axis AXb (AXc).

[Modification 1]
In the first modification, each of the acousto-optic modulation elements AM1, AM3 (AM2, AM4) in the beam switching unit BD1B (BD2B) is rotated 90° around the optical axis AXs from the state shown in Fig. 6. Fig. 12A is a diagram showing an optical path in which mirrors M30 and M32 are added to the first-stage acousto-optic modulation element AM3, lens GK3, and epi-illumination mirror IM3 in Fig. 6, as viewed in the XY plane of the orthogonal coordinate system XYZ, and Fig. 12B is a diagram showing the arrangement of spot lights SPa, SPb, and SPc projected onto the sheet substrate P by the optical path in Fig. 12A.

In this modified example, as shown in FIG. 12B, three spot lights SPa, SPb, and SPc are arranged so as to be aligned at a predetermined interval in the sub-scanning direction (Xt direction). Here, as an example, the drawing line SLnb (n = 1 to 4) by the slot-shaped spot light SPb inclined at -45 degrees is set to a position shifted by an interval ΔXb equivalent to 5 lines in the -Xt direction with respect to the drawing line SLna (n = 1 to 4) by the circular spot light SPa, and the drawing line SLnc (n = 1 to 4) by the slot-shaped spot light SPc inclined at +45 degrees is set to a position shifted by an interval ΔXc equivalent to 5 lines in the +Xt direction. Also, as described above in FIG. 10, the interval in the Xt direction of the drawing lines by the spot lights scanned by each of the adjacent reflecting surfaces of the polygon mirror PM is set to 1/2 of the dimension Xpi in the Xt direction of the pixel Pic.

In order to arrange the three spot lights SPa, SPb, and SPc in the Xt direction, as shown in Fig. 12A, the first-stage acousto-optical modulation element AM3 (as well as the latter-stage acousto-optical modulation element AM1) shown in Fig. 6 is rotated 90 degrees around the optical axis AXs to set the diffraction direction of the acousto-optical modulation element AM3 (AM1) to the -Y direction in the XY plane. Furthermore, each of the three beams LB1a, LB1b, and LB1c (parallel light beams) that intersect at position Pe (see Figs. 7A and 7B) in the crystal of the acousto-optical modulation element AM3 (AM1) is also made to enter the acousto-optical modulation element AM3 (AM1) along a plane parallel to the XZ plane. For this purpose, for example, an image rotator that rotates the optical paths of the three beams LB1a, LB1b, and LB1c by 90 degrees around the optical axis (beam LB1a) is provided in the optical path from the parallel plate 12A and the wedge prisms 12B and 12C shown in FIG. 4 to the first-stage acousto-optical modulation element AM3.

As a result, from the acousto-optical modulation element AM3 in the on state, the zeroth-order beams B3ao, B3bo, and B3co (each a parallel light beam) of the incident beams LB1a, LB1b, and LB1c, and the beams B3a, B3b, and B3c (each a parallel light beam) as the first-order diffracted beams deflected at a predetermined diffraction angle in the -Y direction are emitted. As in Figures 7A and 7B, each of the zeroth-order beams B3ao, B3bo, and B3co and each of the beams B3a, B3b, and B3c are focused by the lens GK3 so as to form a beam waist on the plane Pso where the reflecting surface of the incident mirror IM3 is located. As shown in Figure 12A, in this modified example, the reflecting surface of the incident mirror IM3 is installed so as to form 45 degrees with respect to the XZ plane and the YZ plane, and each of the beams B3a, B3b, and B3c is reflected in the -Y direction.

The central rays of each of the beams B3a, B3b, and B3c that pass through the lens GK3 are parallel to the optical axis AXs, and the beams B3a, B3b, and B3c (respectively divergent beams) reflected by the incident light mirror IM3 are projected onto the mirror M30 in a state where they overlap in the Z direction in FIG. 12A, and the optical path is bent by 90 degrees in the -X direction by the mirror M30. The beams B3a, B3b, and B3c reflected by the mirror M30 are reflected in the -Z direction by the mirror M32, which has a reflecting surface inclined at 45 degrees with respect to the XY and YZ planes. The central rays of each of the beams B3a, B3b, and B3c immediately after being reflected by the mirror M32 are lined up at a predetermined interval in the X direction when viewed in the XY plane.

The beam B3a reflected by the mirror M32 is coaxial with the optical axis AXu3 (or the optical axis AXun shown in FIG. 8) shown in FIG. 7A and FIG. 7B, and enters the lens GK6 as in FIG. 8. Each of the beams B3b and B3c reflected by the mirror M32 enters the lens GK6 shown in FIG. 8 with its central ray symmetrically separated in the X direction across the optical axis AXu3 (or the optical axis AXun shown in FIG. 8) shown in FIG. 7A and FIG. 7B. Therefore, for each of the beams B3a, B3b, and B3c that pass through the lens LGa shown in FIG. 8 and are focused on the surface OPa, the spot SPa' of the beam B3a is located on the optical axis AXun, the spot SPb' of the beam B3b is located a predetermined distance away from the optical axis AXun in the +Z direction, and the spot SPc' of the beam B3c is located a predetermined distance away from the optical axis AXun in the -Z direction.

With the configuration as shown in FIG. 12A, the spot lights SPa, SPb, and SPc projected onto the sheet substrate P can be arranged in the sub-scanning direction (Xt direction) as shown in FIG. 12B. In this modified example, as shown in FIG. 12B, the projection positions of the slot-shaped (elliptical) spot lights SPb and SPc are shifted by intervals ΔXb and ΔXc across multiple (here, five) drawing lines in the Xt direction from the projection position of the circular spot light SPa. Therefore, naturally, the data strings in the drawing signal corresponding to the diagonal line pattern or diagonal edge portion drawn by each of the spot lights SPb and SPc (data strings for generating each of the drawing signals SDb and SDc shown in FIG. 11) are shifted in the sub-scanning direction (Xt direction) by an amount equivalent to the intervals ΔXb and ΔXc with respect to the data string corresponding to the pattern drawn by the spot light SPa (data strings for generating the drawing signal SDa shown in FIG. 11) and stored.

[Modification 2]
In the beam compression system OM2 shown in Figures 5A to 5C, the cross-sectional shapes (circular) of the incident beams LB1b and LB1c are compressed one-dimensionally using two cylindrical lenses 10G3 and 10G4, but other optical elements may be used. Figure 13 is a diagram showing a schematic configuration of the beam compression system OM2 according to this modification. In this modification, the beam compression system OM2 is configured by an optical element 10G3', such as a one-dimensional micro prism array or a one-dimensional Fresnel lens, on which the beam LB1b (LB1c) with a circular cross section expanded by the beam expansion system OM1 in Figures 5A to 5C is incident, and a cylindrical lens 10G4' having negative power. When optical element 10G3' is made into a one-dimensional microprism array, when viewed within the plane of FIG. 13, it is configured such that multiple prism sections, each having a fine wedge-shaped cross section and extending one-dimensionally in a direction perpendicular to the plane of the paper, are arranged symmetrically on either side of the optical axis AXb (AXc), and the apex angle of the wedge of the prism section becomes larger as it moves away from the optical axis AXb (AXc).

As a result, in the cross section of the incident beam LB1b (LB1c), the refraction angle at the prism portion increases as the beam moves away from the optical axis AXb (AXc) of the optical element 10G3', and the beam LB1b (LB1c) is compressed (converged) toward the optical axis AXb (AXc). The cylindrical lens 10G4' diverges the compressed (converged) beam LB1b (LB1c) so that it becomes a nearly parallel beam. Also, when viewed in a plane perpendicular to the paper surface of FIG. 13 and including the optical axis AXb (AXc), neither the optical element 10G3' nor the cylindrical lens 10G4' has power (refractive power), so the incident beam LB1b (LB1c) proceeds as a parallel beam.

As a result, the shape of the intensity distribution in the cross section of the beam LB1b (LB1c) emerging from the cylindrical lens 10G4' becomes slot-shaped (elliptical). In addition, when a one-dimensional Fresnel lens is used as the optical element 10G3', the cross-sectional shape of the beam LB1b (LB1c) emerging from the cylindrical lens 10G4' can be made slot-shaped (elliptical) by substantially the same action. In this modified example, the major axis direction of each of the slot-shaped (elliptical) spot lights SPb, SPc projected onto the sheet substrate P can be tilted with respect to the main scanning direction (Yt direction) by rotating the entire beam compression system OM2 consisting of the optical element 10G3' and the cylindrical lens 10G4' around the optical axis AXb (AXc). In addition, the cylindrical lens 10G4' may be a one-dimensional Fresnel lens having negative power (refractive power). Furthermore, the entrance surface (or exit surface) of the cylindrical lens 10G4' does not have to be a perfect cylindrical surface, and may be an approximate cylindrical surface that is approximated by a high-order function, like an aspherical lens.

Second Embodiment
In the first embodiment and modified examples described above, the optical paths were set in the beam combining units BD1A and BD2A as shown in FIG. 4 so that the beam LB1a (LB2a) from the light source device LS1A (LS2A) for generating the circular spot light SPa, the beam LB1b (LB2b) from the light source device LS1B (LS2B) for generating the slot-shaped spot light SPb, and the beam LB1c (LB2c) from the light source device LS1C (LS2C) for generating the slot-shaped spot light SPc intersect within the crystals of the acousto-optical modulation elements AM1 to AM4 in the beam switching units BD1B and BD2B. In this embodiment, the three beams LB1a (LB2a), LB1b (LB2b), and LB1c (LB2c) incident on the acousto-optical modulation elements AM1 to AM4 are combined coaxially with the optical axis AXs (see Figure 6) in the beam switching unit BD1B (BD2B).

Figure 14 is a diagram showing a configuration according to a second embodiment in which beams LB1a, LB1b, and LB1c from each of the beam shape transformation units 10A, 10B, and 10C shown in Figure 4 are coaxially synthesized. In this embodiment, the coaxial synthesis is performed by a configuration in which the polarization state of each of the beams LB1a, LB1b, and LB1c is switched at high speed by an electro-optical element. In Figure 14, the P-polarized beam LB1a from the light source device LS1A that enters the beam shape transformation unit 10A enters the first surface of the polarizing beam splitter BS1 via the lenses 10G5 and 10G6 of the beam reduction system OM3. The P-polarized beam LB1b from the light source device LS1B that enters the beam shape transformation unit 10B is converted to S-polarized light via the lenses 10G5 and 10G6 of the beam reduction system OM3 and the 1/2 wavelength plate HWP, and enters the second surface perpendicular to the first surface of the polarizing beam splitter BS1.

The P-polarized beam LB1a incident on the first surface of the polarizing beam splitter BS1 passes through the polarization separation surface of the polarizing beam splitter BS1 and enters the first electro-optical element EOa. The electro-optical element EOa switches the direction of linear polarization of the incident beam by turning on/off the drive signal SSa, which applies an electric field to the internal crystal. Therefore, when the drive signal SSa is off, the electro-optical element EOa transmits the P-polarized beam LB1a that has passed through the polarizing beam splitter BS1 as is, and it enters the second polarizing beam splitter BS2. The second polarizing beam splitter BS2 is also arranged to transmit P-polarized light and reflect S-polarized light, so the P-polarized beam LB1a passes through the polarizing beam splitter BS2 and enters the second electro-optical element EOb.

The second electro-optical element EOb is the same as the first electro-optical element EOa, and switches the polarization state of the incident beam by turning the drive signal SSb On/Off. When the drive signal SSb is Off, the electro-optical element EOb transmits the P-polarized beam LB1a from the polarizing beam splitter BS2 as is, and makes it incident on the third polarizing beam splitter BS3. The third polarizing beam splitter BS3 is also arranged to transmit P-polarized light and reflect S-polarized light, so that the P-polarized beam LB1a transmits through the polarizing beam splitter BS3 and proceeds coaxially with the optical axis AXs in the subsequent beam switching unit BD1B (BD2B).

On the other hand, beam LB1b, which becomes S-polarized from beam shape deformation unit 10B and enters polarizing beam splitter BS1, is reflected by the polarization separation surface of polarizing beam splitter BS1 and enters electro-optical element EOa. When drive signal SSa is Off, S-polarized beam LB1b passes through electro-optical element EOa as is and enters polarizing beam splitter BS2 at the same entrance surface as beam LB1a. Almost all of the S-polarized beam LB1b that enters polarizing beam splitter BS2 is reflected and enters beam trap TRa where it is absorbed.

Furthermore, the P-polarized beam LB1c from the light source device LS1C that enters the beam shape transformation unit 10C is converted to S-polarized light via the lenses 10G5 and 10G6 and the 1/2 wavelength plate HWP of the beam reduction system OM3, is reflected at a right angle by the mirror M40A, and enters the second surface (the surface opposite the beam trap TRa) of the polarizing beam splitter BS2. Since the polarizing beam splitter BS2 reflects S-polarized light, the S-polarized beam LB1c is reflected so as to have a coaxial optical path with the other beams LB1a and LB1b, and enters the electro-optical element EOb. When the drive signal SSb is Off, the electro-optical element EOb transmits the S-polarized beam LB1c from the polarizing beam splitter BS2 as it is, and causes it to enter the third polarizing beam splitter BS3. The third polarizing beam splitter BS3 is also arranged to reflect S-polarized light, so that the S-polarized beam LB1c is reflected by the polarizing beam splitter BS3 and absorbed in the beam trap TRb.

As described above, when the drive signals SSa and SSb applied to the two electro-optical elements EOa and EOb arranged in series are both in the Off state, only the P-polarized beam LB1a from the light source device LS1A is emitted coaxially with the optical axis AXs from the polarizing beam splitter BS3. Next, a case will be described in which the drive signal SSa applied to the first electro-optical element EOa is in the On state and the drive signal SSb applied to the second electro-optical element EOb is in the Off state. In that case, the P-polarized beam LB1a incident on the first electro-optical element EOa via the polarizing beam splitter BS1 is switched to S-polarized light. Therefore, the S-polarized beam LB1a is reflected by the second polarizing beam splitter BS2 and absorbed by the beam trap TRa.

On the other hand, the S-polarized beam LB1b reflected by the polarizing beam splitter BS1 and incident on the first electro-optical element EOa is switched to P-polarized light. Therefore, the beam LB1b, which has become P-polarized light, passes through the second polarizing beam splitter BS2, the second electro-optical element EOb, and the third polarizing beam splitter BS3 as it is, and is emitted coaxially with the optical axis AXs. At that time, the beam LB1c, which has become S-polarized light from the beam shape transformation unit 10C, is reflected by the second polarizing beam splitter BS2 via the mirror M40A, passes through the second electro-optical element EOb in the Off state as it is, is reflected by the third polarizing beam splitter BS3, and is absorbed by the beam trap TRb. As described above, when the electro-optical element EOa is in the On state and the electro-optical element EOb is in the Off state, only the P-polarized beam LB1b is emitted from the third polarizing beam splitter BS3 coaxially with the optical axis AXs.

Next, a case will be described in which the drive signals SSa and SSb are applied so that both the first electro-optical element EOa and the second electro-optical element EOb are in the On state. In this case, since the electro-optical element EOa is in the On state, the P-polarized beam LB1a from the beam shape transformation unit 10A passes through the polarized beam splitter BS1 and is converted to S-polarized light through the electro-optical element EOa, so it is reflected by the polarized beam splitter BS1 and absorbed by the beam trap TRa. Also, the S-polarized beam LB1b from the beam shape transformation unit 10B passes through the electro-optical element EOa after being reflected by the polarized beam splitter BS1 and is converted to P-polarized light, so it passes through the next polarized beam splitter BS2. However, the P-polarized beam LB1b that passed through the polarized beam splitter BS2 passes through the second electro-optical element EOb in the On state and is converted to S-polarized light, so it is reflected by the polarized beam splitter BS3 and absorbed by the beam trap TRb.

On the other hand, the S-polarized beam LB1c from the beam shape transformation unit 10C is reflected by the mirror M40A and the polarizing beam splitter BS2, then passes through the electro-optical element EOb in the On state to be converted to P-polarized light, and passes through the next polarizing beam splitter BS3, emerging coaxially with the optical axis AXs. Thus, in this embodiment, the polarizing beam splitters BS1, BS2, BS3 and the electro-optical elements EOa and EOb function as a beam synthesis unit that synthesizes the three beams LB1a, LB1b, and LB1c to travel along the optical axis AXs.

In this embodiment, two electro-optical elements EOa and EOb are used to combine the beams LB1a, LB1b, and LB1c from the three light source devices LS1A, LS1B, and LS1C coaxially in the same linear polarization state. For this reason, the drawing data storage unit 100B in the control device 100 shown in Figure 4 stores information (bitmap information) that sets the on/off of the drive signals SSa and SSb (high-voltage direct current potential) applied to each of the electro-optical elements EOa and EOb in association with the pixel map information of the pattern to be drawn.

As shown in FIG. 14, by combining the polarizing beam splitters BS1, BS2, BS3 and the electro-optical elements EOa and EOb, any one of the beams LB1a, LB1b, and LB1c from each of the light source devices LS1A, LS1B, and LS1C can be made coaxial with the optical axis AXs and incident on the first-stage acousto-optical modulation element AM3 of the beam switching unit BD1B. Furthermore, since an amplitude-splitting type beam splitter that does not have polarization separation characteristics is not used for coaxial synthesis, the attenuation of the light amount of each of the beams LB1a, LB1b, and LB1c can be suppressed and the polarization direction can be made the same P-polarized light. Therefore, the diffraction efficiency for each of the beams LB1a, LB1b, and LB1c passing through the acousto-optical modulation element AM3 (AM1) is the same, and the variation in the light amount (intensity) of each of the spot lights SPa, SPb, and SPc projected on the sheet substrate P is suppressed.

Figure 15 is a diagram illustrating an example of a pattern drawing operation using the beam combining unit BD1A of Figure 14, the beam switching unit BD1B of Figure 6, and the drawing unit MU3 (or MU1). In this embodiment, each of the three beams LB1a, LB1b, and LB1c incident on the beam switching unit BD1B (acousto-optical modulation elements AM3 and AM1) is set coaxially with the optical axis AXs, so that each of the three spot beams SPa, SPb, and SPc is also projected to the same position in the main scanning direction on the drawing line SL3 (SL1).

Figure 15 shows, as an example, a case in which a pattern in which a rectangular pattern portion PT4, a line pattern portion PT5 extending in the Yt direction, and a diagonal line pattern portion PT6 are connected in the Yt direction is drawn by drawing unit MU3. Pattern portion PT4 has a diagonal edge portion E4a that is inclined with respect to the drawing line SL3 drawn by drawing unit MU3, and an edge portion E4b that extends linearly in the Xt direction. Diagonal line pattern portion PT6 has diagonal edge portions E6a and E6b that are inclined in the opposite direction to diagonal edge portion E4a.

When pattern drawing is performed along the drawing line SL3 for a pattern such as that shown in Figure 15, in the region Ar1 on the drawing line SL3 that crosses the diagonal edge portion E4a, pattern drawing (pulse emission of light source device LS1B) is performed in accordance with the pixel bit information of the drawing signal SDb and the clock signal CLK (SDb∩CLK) by the slot-shaped (oval-shaped) spot light SPb inclined at -45 degrees, as in Figure 10 above. In the region Ar2 on the drawing line SL3 including the edge portion E4b of the pattern portion PT4, pattern drawing (pulse emission of light source device LS1A) is performed by the circular spot light SPa in accordance with the pixel bit information of the drawing signal SDa and the clock signal CLK (SDa∩CLK). Furthermore, in the area Ar3 on the drawing line SL3 including the diagonal edge portions E6a and E6b of the pattern portion PT6, a slot-shaped (oblong) spot light SPc inclined at +45 degrees is used to draw a pattern (pulse emission of the light source device LS1C) in accordance with the pixel bit information of the drawing signal SDc and the clock signal CLK (SDc∩CLK).

In the case of drawing such a pattern, with regard to the drive signals SSa and SSb applied to each of the electro-optical elements EOa and EOb in FIG. 14, only the drive signal SSa is in the On state from time Ts1, just before the spot light SPb is projected in area Ar1, to time Ts2 when the projection of the spot light SPb is switched on, and both the drive signals SSa and SSb are in the On state from time Ts3, just before the spot light SPc is projected in area Ar3, to time Ts4 when the drawing of the diagonal pattern portion PT6 in area Ar3 is completed.

In addition, the upper limit of the switching response frequency of the electro-optical elements EOa and EOb is often lower than the 400 MHz (period 2.5 nS) pulse emission frequency of the light source devices LS1B and LS1C. For this reason, a time width (Ts2-Ts1, or Ts4-Ts3) is set that is equal to or greater than the minimum time width that can turn the drive signals SSa and SSb to the On state, and the regions Ar1 and Ar3 (diagonal edge portions E4a, E6a, and E6b) are set within that time width.

As described above, when drawing a pattern including a narrow diagonal line pattern portion PT6 and a diagonal edge portion E4a, the circular spot light SPa and the inclined slot-shaped (elliptical) spot lights SPb and SPc are set to be projected alternatively at the same position in the main scanning direction. Therefore, as explained in FIG. 11, there is no need to shift each data string for the spot lights SPa, SPb, and SPc of the pixel bit information of the numerous pixels Pic lined up along the drawing line SLn by the number of pixels corresponding to the intervals ΔYb and ΔYc. This reduces the effort required to generate each data string.

[Modification 3]
As in the first embodiment, when a plurality of spot lights SPa, SPb, and SPc can be simultaneously scanned on the sheet substrate P by the rotation of the polygon mirror PM, it is possible to easily accommodate differences in sensitivity and thickness of the photosensitive layer (photoresist layer) formed on the surface of the sheet substrate P. For example, in an exposure device that performs pattern drawing using only a single spot light SPa (circular), the beam LB1a emitted from the light source device LS1A is adjusted to be as bright as possible in order to shorten the exposure processing time. For this reason, the usable photoresist has a recommended sensitivity range and a recommended thickness range corresponding to the light intensity of the beam LB1a. If the sensitivity of the photoresist layer used is considerably low or is significantly thicker than the recommended thickness, it is difficult to increase the light intensity of the beam LB1a (spot light SPa), so that the rotation speed of the polygon mirror PM (scanning speed of the spot light SPa) and the movement speed of the sheet substrate P in the sub-scanning direction are greatly reduced.

That is, the supplyable exposure amount (supplied dose amount) determined by the light intensity of the spot light SPa and the state of scanning exposure (speed, etc.) is adjusted to match the required exposure amount (required dose amount) determined by the sensitivity and thickness of the resist layer formed on the sheet substrate P. In the first embodiment, each of the drawing units MU1 to MU4 can selectively project one, two, or all of the three spot lights SPa, SPb, and SPc, so that the adjustment range of the supplied dose amount can be greatly expanded. Furthermore, each of the spot lights SPa, SPb, and SPc can be pulse-projected on the sheet substrate P at high speed in pixel Pic units in the drawing data, so that a special exposure method that gives a larger dose amount than usual to pixels corresponding to the edge portion of the pattern and the pixels adjacent to it is also possible.

16 is a diagram showing an example of a special exposure drawing operation that increases the amount of exposure given to each peripheral edge portion of a plurality of rectangular patterns arranged in a matrix using an exposure device according to the first embodiment or modified example 1. The special exposure method can be used when the resist layer formed on the sheet substrate P is negative type and has a thickness several to ten times that of a general thickness (0.8 μm to 2 μm), as disclosed in International Publication No. 2019/049940, for example.

16, when one pixel Pic on the drawing data is set to 2 μm square on the sheet substrate P, the rectangular pattern PT7 is defined as having a size of 9 pixels (18 μm) in the Yt direction and 11 pixels (22 μm) in the Xt direction, and is arranged in a matrix shape with an interval of 3 pixels (6 μm) in each of the Xt and Yt directions. Each rectangular pattern PT7 is composed of a peripheral edge portion PT7a made of pixels linearly arranged in each of the Xt and Yt directions, and a rectangular pattern portion PT7b made of 7 pixels x 9 pixels inside the peripheral edge portion PT7a. Assume that the exposure device according to the first embodiment scans spot lights SPa, SPb, and SPc along drawing lines SL3a and SL3b by the drawing unit MU3, for example.

In this modified example, each pixel (ON pixel) constituting the inner rectangular pattern portion PT7b is exposed to two pulses of circular spot light SPa in each of the Xt and Yt directions, and each pixel (ON pixel) constituting the peripheral edge portion PT7a is additionally exposed to two pulses of slot-shaped (oblong) spot light SPb, SPc in each of the Xt and Yt directions along with two pulses of circular spot light SPa. Therefore, on the drawing line SL3a, 18 pulses of circular spot light SPa corresponding to the full width (9 pixels) of the rectangular pattern PT7 in the Xt direction are irradiated based on the drawing signal SDa and the clock signal CLK.

In addition, on the drawing line SL3a, in order to additionally expose the first pixel constituting the peripheral edge portion PT7a on the +Yt direction side of the rectangular pattern PT7 and the ninth pixel constituting the peripheral edge portion PT7a on the -Yt direction side, two pulses of spot light SPb, SPc are irradiated to each of the first pixel and the ninth pixel based on the drawing signals SDb, SDc and the clock signal CLK. Note that in the first embodiment, the three spot lights SPa, SPb, SPc are shifted in the Yt direction at intervals ΔYb, ΔYc as described in FIG. 10 above, so in this modified example, as described in FIG. 11, the data string of pixel bit information corresponding to each of the drawing signals SDa, SDb, SDc has the information position (bit position) shifted by the number of pixels according to the intervals ΔYb, ΔYc.

In addition, on the drawing line SL3b, additional exposure is performed on a row of nine pixels lined up in the Yt direction to form the peripheral edge portion PT7a on the Xt direction side of the rectangular pattern PT7. Therefore, on the drawing line SL3b, two pulses of spot light SPa, SPb, SPc are irradiated onto each of the nine pixels from the first pixel constituting the peripheral edge portion PT7a on the +Yt direction side of the rectangular pattern PT7 to the ninth pixel constituting the peripheral edge portion PT7a on the -Yt direction side, based on the drawing signals SDa, SDb, SDc and the clock signal CLK.

In this modified example, the pixel to be additionally exposed (ON pixel) is successively irradiated with a circular spot light SPa, a slot-shaped spot light SPb inclined at -45 degrees, and a slot-shaped spot light SPc inclined at +45 degrees. Therefore, when the light intensity of each spot light is the same, the additionally exposed pixel is given a maximum exposure amount of about three times. However, the amount of exposure required for additional exposure may be about 1.5 times or 2 times. In that case, a light-attenuating member capable of variably adjusting the beam intensity may be provided in the optical path of the beams LB1b and LB1c emitted from the light source devices LS1B and LS1C so that the light intensity of each of the slot-shaped spot lights SPb and SPc used for additional exposure is attenuated to about 25% and 50% of the light intensity of the spot light SPa. As such a light-attenuating member, a configuration in which the beam LB1b (LB1c) from the light source device LS1B (LS1C) passes through a rotatable 1/2 wavelength plate and a polarizing beam splitter in that order is preferable. In this case, by rotating the half-wave plate around the central ray of the beam, the intensity of the beam reflected (or transmitted) by the polarizing beam splitter can be continuously adjusted, for example, in the range from 10% to 90%.

In this modified example, during additional exposure, the pixels constituting the peripheral edge portion PT7a are irradiated with overlapping slot-shaped spot light SPb inclined at -45 degrees and slot-shaped spot light SPc inclined at +45 degrees, so the light intensity distribution due to the overlapping of the two spot lights SPb and SPc becomes close to a rectangle with rounded corners. Therefore, if the entire device formation area to be exposed on the sheet substrate P includes only the peripheral edge portion PT7a extending in the Xt and Yt directions as shown in Figure 16 and does not include a diagonal edge portion or a diagonal line pattern, the pixels (ON pixels) of the peripheral edge portion PT7a may be exposed only with the overlapping of the two spot lights SPb and SPc.

[Modification 4]
In the above first embodiment, second embodiment, and each modified example, three beams Bna, Bnb, Bnc (n=1 to 4) generated from each of three light source devices LS1A, LS1B, LS1C (LS2A, LS2B, LS2C) are switched and supplied to one of a plurality of imaging units MUn by the acousto-optical modulation elements AM1, AM3 (AM2, AM4) arranged in series in the beam switching unit BD1B (BD2B). However, a configuration may be used in which the imaging beams from each of the two light source devices are supplied directly to one imaging unit without passing through the beam switching unit BD1B (BD2B).

Figure 17 is a diagram showing a schematic configuration of a beam combining unit according to modified example 4 using only two light source devices. In Figure 17, the same members as those shown in Figures 1 and 4 above are given the same reference numerals, and detailed descriptions thereof are omitted. In this modified example, beams LB1b and LB1c from the two light source devices LS1B and LS1C are reflected by mirrors M50 and M52, respectively, and enter beam shape transformation units 10B and 10C shown in Figure 4 (as well as Figures 5A to 5C). In Figure 17, beam LB1b from light source device LS1B is emitted in the +X direction parallel to the XY plane of the orthogonal coordinate system XYZ, and beam LB1c from light source device LS1C is emitted in the -X direction parallel to the XY plane in a positional relationship approximately coaxial with beam LB1b.

The entire beam compression system OM2 (see Figures 5A to 5C) provided in the beam shape transformation unit 10B is rotatable around the central ray (optical axis AXb in Figures 5A to 5C) of the incident beam LB1b. The beam compression system OM2 is set to rotate by 45° by a drive mechanism 20B including actuators such as a motor and an air piston. Therefore, the beam LB1b emitted in the +Y direction from the beam shape transformation unit 10B is converted into a parallel beam whose cross-sectional distribution in the YZ plane is slot-shaped (oblong). Similarly, the entire beam compression system OM2 (see Figures 5A to 5C) provided in the beam shape transformation unit 10C is rotatable around the central ray (optical axis AXc in Figures 5A to 5C) of the incident beam LB1c, and is set to rotate by 45° by a drive mechanism 20C including actuators such as a motor and an air piston. Therefore, the beam LB1c emerging from the beam shape modifying section 10C in the +Y direction is converted into a parallel light beam whose cross-sectional distribution in the YZ plane is slot-shaped (elliptical).

Beam LB1b from beam shape transformation unit 10B is reflected in the +X direction by mirror M51 and then projected onto one reflecting surface of V-shaped mirror M54. Similarly, beam LB1c from beam shape transformation unit 10C is reflected in the -X direction by mirror M53 and then projected onto the other reflecting surface of V-shaped mirror M54. One reflecting surface and the other reflecting surface of V-shaped mirror M54 are set to intersect at a predetermined angle to form a ridgeline parallel to the Z axis in FIG. 17. The ridgeline of V-shaped mirror M54 (parallel to the Z axis) is set to be perpendicular to the optical axis AXu1 (see FIG. 2) of drawing unit MU1, for example, when viewed in the XY plane in FIG. 17.

Beam LB1b reflected by one reflecting surface of V-shaped mirror M54 and beam LB1c reflected by the other reflecting surface of V-shaped mirror M54 proceed in the +Y direction parallel to optical axis AXu1 in the XY plane and symmetrically approaching each other on either side of optical axis AXu1, and enter prism block 22. Prism block 22 has a ridgeline extending parallel to the Z axis so as to be perpendicular to optical axis AXu1, and refracts (deflects) each of the incident beams LB1b and LB1c toward optical axis AXu1 by a predetermined angle (for example, 1° or less). After passing through prism block 22, two beams LB1b and LB1c cross each other on plane Pe' (corresponding to plane Pe' in FIG. 8) perpendicular to optical axis AXu1, and then proceed to enter drawing unit MU1 while expanding.

With the above configuration, only the two spot lights SPb, SPc by each of the beams LB1b, LB1c projected from the drawing unit MU1 are focused on the sheet substrate P. In this modified example, each of the two spot lights SPb, SPc has a slot-shaped (elliptical) intensity distribution, and the major axis direction of the distribution can be changed by an angle of 45° on the sheet substrate P by the driving mechanisms 20B, 20C. The amount of change in the rotation angle of the beam compression system OM2 in the beam shape transformation units 10B, 10C by the driving mechanisms 20B, 20C may be, for example, in 12 steps of 15° within the range of ±90°, or may be set to any angle without any steps.

In this modified example, when a pattern is drawn using only two slot-shaped (elliptical) spot lights SPb and SPc, the drawing unit MU1 determines in advance what pattern portion (straight edge portion extending in the main scanning direction or sub-scanning direction, or diagonal edge portion, etc.) is included in the pattern to be drawn and at what position in the sub-scanning direction on the sheet substrate P, and accordingly, the long axis direction of each of the spot lights SPb and SPc is rotated in advance (just before drawing the corresponding edge portion) by the drive mechanisms 20B and 20C. As described above in FIG. 9A, in the wiring patterns in electronic devices, 45-degree diagonal line patterns (PT2 in FIG. 9A) are often used, but in addition, line patterns and pattern edge portions inclined at about 30 degrees or 60 degrees to the main scanning direction or sub-scanning direction are also used. Taking such cases into consideration, it is desirable to tilt the major axis direction of each of the spot lights SPb, SPc having a slot-shaped (elliptical) intensity distribution within the range of +25 degrees to 65 degrees, or within the range of -25 degrees to 65 degrees, relative to the main scanning direction or sub-scanning direction.

Figures 18A to 18D are schematic diagrams showing combinations of the major axis directions of the spot lights SPb and SPc that are switched according to the directionality of the edge portion of the pattern. Figure 18A shows a case where the centers of the spot light SPb tilted at -45 degrees and the spot light SPc tilted at +45 degrees are located in the Yt (Y) direction on one drawing line SLn at an interval (ΔYb + ΔYc) as in the state described in FIG. 10 above. When the pattern drawn by the drawing unit MU1 is composed only of straight edge portions extending in the Xt and Yt directions as described in FIG. 16 above, the pulse emission of the beams LB1b and LB1c from the light source devices LS1B and LS1C is controlled so that the spot lights SPb and SPc tilted at ±45 degrees are projected overlappingly onto the ON pixels to which the spot lights should be projected among all the pixels Pic on the drawing data. As shown in FIG. 18A, the initial state is a state in which the spot light SPb is inclined at −45 degrees and the spot light SPc is inclined at +45 degrees.

Figure 18B shows a state in which only the spot light SPc has been rotated 90 degrees counterclockwise from the initial tilt by the drive mechanism 20C, and Figure 18C shows a state in which only the spot light SPb has been rotated 90 degrees clockwise from the initial tilt by the drive mechanism 20B. As in Figure 18B or 18C, by tilting the two spot lights SPb and SPc in the same direction and projecting the spot lights SPb and SPc in an overlapping manner onto the on-pixels (Pic) of the diagonal edge portion or diagonal line pattern, the exposure amount of the pixels of the edge portion can be increased.

Figure 18D also shows a state in which the driving mechanisms 20B and 20C rotate the spot light SPb 45 degrees clockwise from the initial tilt, and rotate the spot light SPc 45 degrees counterclockwise from the initial tilt. In the case of Figure 18D, the two spot lights SPb and SPc are both set so that their major axis directions are perpendicular to the drawing line SLn. The setting of Figure 18D is particularly suitable for drawing linear line patterns that extend parallel to or perpendicular to the drawing line SLn.

As described above, the change (switching) operation of the long axis direction of the spot lights SPb, SPc by rotating the beam compression system OM2 by the drive mechanisms 20B, 20C cannot be performed during one scan of the spot lights SPb, SPc along the drawing line SLn. Therefore, if necessary, the switching operation is performed at a timing when the time required for the switching operation of the spot lights SPb, SPc in the long axis direction becomes shorter than the time it takes for the sheet substrate P to move through an area where no pattern drawing is performed at all (an area where off pixels where the spot lights SPb, SPc are not pulsed) is.

[Modification 5]
19 is a schematic diagram showing an optical configuration in which two slot-shaped (long elliptical) spot lights SPb and SPc are produced by a beam LBe from one light source device LSe. For convenience of explanation, the beam LBe from the light source device LSe is assumed to be emitted parallel to the X-axis of the orthogonal coordinate system XYZ. The light source device LSe is a fiber amplifier laser light source (oscillation of ultraviolet pulse light with a wavelength of 355 nm at a frequency of 400 MHz) similar to the light source devices LS1B and LS1C described above. The P-polarized beam LBe (parallel light beam having a circular cross-sectional distribution with a diameter of about 0.5 to 1 mm) from the light source device LSe is incident on an electro-optical element EOc similar to the electro-optical elements EOa and EOb described above in FIG. 14. When the drive signal SSc (high voltage DC potential) is applied (when in the on state), the electro-optical element EOc converts the incident beam LBe from P polarized light to S polarized light and emits it, and when the drive signal SSc is not applied (when in the off state), it emits the P polarized beam LBe as is.

The beam LBe from the electro-optical element EOc is bent at a right angle by the mirror M55 so that it travels in the +Y direction parallel to the Y axis, and then enters the beam expansion system OM1 (see Figures 5A to 5C). The beam expansion system OM1 expands the diameter of the incident beam LBe by about 10 times to form a parallel beam and emits it toward the polarizing beam splitter BS4. The polarizing beam splitter BS4, which acts as a light splitting unit, transmits the beam LBe as is when the beam LBe is P-polarized, and reflects the beam LBe at a right angle when the beam LBe is S-polarized. Here, if the P-polarized beam LBe that has passed through the polarizing beam splitter BS4 is called beam LB1c, and the S-polarized beam LBe that has been reflected by the polarizing beam splitter BS4 is called beam LB1b, the split beams LB1c and LB1b each travel along a different optical path.

Beam LB1b (S-polarized) from polarizing beam splitter BS4 passes through half-wave plate HWP and is converted to P-polarized light with its polarization direction rotated by 90 degrees, and then enters beam compression system OM2b, which functions as a beam shape modification unit. Similarly, beam LB1c (P-polarized) from polarizing beam splitter BS4 passes through half-wave plate HWP and is converted to S-polarized light with its polarization direction rotated by 90 degrees, and then enters beam compression system OM2c, which functions as a beam shape modification unit. Each of beam compression systems OM2b and OM2c is similarly configured using the optical members shown in Figures 5A to 5C or 13, and is installed so as to form a relative angle of 90 degrees around the central ray (optical axis) of each beam LB1b and LB1c.

After passing through the beam compression system OM2b, the beam LB1b is reflected at a right angle by the mirror M57 so that it becomes parallel to the Y axis, travels in the +Y direction, and enters the polarizing beam splitter BS5. Also, the beam LB1c that travels in the -X direction through the beam compression system OM2c is entered into the polarizing beam splitter BS5. The polarizing beam splitter BS5, which functions as a beam combining section, is arranged to transmit the beam LB1b that has become P polarized and reflect the beam LB1c that has become S polarized, and to combine the beams LB1b and LB1c coaxially. The beams LB1b and LB1c that are emitted in the +Y direction from the polarizing beam splitter BS5 are converted into parallel light beams with the beam diameter reduced to about 1/10 by the beam reduction system OM3, which has the same configuration as that shown in Figures 5A to 5C.

Beams LB1b and LB1c that have passed through the beam reduction system OM3 pass through an amplitude division type beam splitter BS6 to become beams B1b and B1c, respectively, and enter the drawing unit MU1 so as to be coaxial with the optical axis AXu1 of the drawing unit MU1. Also, beam LB1a (a parallel light beam having a circular cross-sectional distribution with a diameter of about 0.5 to 1 mm) traveling in the -X direction from light source device LS1A, which is made with the same characteristics as light source device LSe, is reflected by beam splitter BS6 to become beam B1a, which enters drawing unit MU1 so as to be coaxial with the optical axis AXu1 of the drawing unit MU1. Beam splitter BS6 is arranged to combine the three beams B1a, B1b, and B1c coaxially, but about half of the light amount of each of the incident beams B1a, B1b, and B1c is absorbed by beam trap TRc arranged on the -X side of beam splitter BS6.

In this modified example, when the electro-optical element EOc is in the off state and the pixel bit value of the drawing signal SDc supplied to the light source device LSe is "1", the light source device LSe emits a pulsed beam LBe. Since the electro-optical element EOc is in the off state, the beam LBe (P polarized light) passes through the polarized beam splitter BS4, passes through the half-wave plate HWP and is converted to S polarized light, passes through the beam compression system OM2c and is reflected by the polarized beam splitter BS5, passes through the beam reduction system OM3 and the beam splitter BS6, and is supplied to the drawing unit MU1 as the beam B1c that generates the spot light SPc. When the electro-optical element EOc is in the on state and the pixel bit value of the drawing signal SDb supplied to the light source device LSe is "1", the light source device LSe emits a pulsed beam LBe. Since the electro-optical element EOc is in the on state, the P-polarized beam LBe is converted to S-polarized light, reflected by the polarizing beam splitter BS4, converted to P-polarized light through the half-wave plate HWP, passes through the beam compression system OM2b, transmits through the polarizing beam splitter BS5, and passes through the beam reduction system OM3 and the beam splitter BS6 to become the beam B1b that generates the spot light SPb, and is supplied to the drawing unit MU1.

In this modified example, the electro-optical element EOc is used to switch between the spot light SPb and the spot light SPc. For this reason, the drawing data storage unit 100B in the control device 100 shown in FIG. 4 stores information (bitmap information) that sets the on/off of the drive signal SSc applied to the electro-optical element EOc in association with the pixel map information of the pattern to be drawn. As a result, the on/off of the electro-optical element EOc complementarily switches the linear polarization direction of the beam LB1b emitted from the beam compression system OM2b as the beam shape transformation unit and the linear polarization direction of the beam LB1c emitted from the beam compression system OM2c as the beam shape transformation unit.

In this modified example, the circular spot light SPa produced by the beam B1a from the light source device LS1A can be constantly projected as pulsed light onto the sheet substrate P in response to the pixel bit value "1" of the drawing signal SDa and the clock signal CLK. On the other hand, the spot light SPb produced by the beam B1b generated from the beam LBe from the light source device LSe and the spot light SPc produced by the beam B1c are projected as pulsed light onto the sheet substrate P in response to the pixel bit value "1" of the drawing signals SDb, SDc and the clock signal CLK by switching the electro-optical element EOc between the off state and the on state.

Therefore, in this modified example, the edge pixels constituting the diagonal edge portion or diagonal line pattern can be exposed not only to the slot-shaped (elliptical) spot light SPb or SPc but also to the circular spot light SPa in a superimposed manner. In this case, the effect of increasing the exposure amount of the edge pixels while reducing the jaggedness of the edge portion can be obtained.

[Modification 6]
Fig. 20 is a diagram showing a schematic configuration of a modified example of a light source device and a beam combining unit BD1A (BD2A) including beam shape transforming units 10B and 10C. In this modified example, a circular spot light SPa and two slot-shaped (oblong) spot light SPb and SPc are generated by the beam LBe from the light source device LSe shown in Fig. 19 and the beam LB1a from the light source device LS1A shown in Fig. 4, and the beam switching unit BD1B (BD2B) shown in Fig. 6 is also used to enable pattern exposure by each of the multiple drawing units MU1 to MU4.

In FIG. 20, the ultraviolet wavelength region beam LBe (a parallel light beam with a circular cross section with a diameter of about 0.5 to 1 mm) pulsed from the light source device LSe at a frequency of 400 MHz is incident on the acousto-optical modulation element AM5 under the condition of Bragg diffraction. The acousto-optical modulation element AM5 is switched between an on state (diffracted light generating state) and an off state by the drive signal SSe. When the acousto-optical modulation element AM5 is in the off state, the incident beam LBe passes through as it is and is incident on the lens GK7 coaxially with the optical axis. The acousto-optical modulation element AM5 is disposed at the front focal position of the lens GK7, and the epi-reflection mirror IM5 is disposed at the rear focal position of the lens GK7. Therefore, the respective arrangements of the acousto-optical modulation element AM5, lens GK7, and epi-reflection mirror IM5 in FIG. 20 are the same as the respective arrangements of the acousto-optical modulation element AM3, lens GK3, and epi-reflection mirror IM3 described above in FIG. 7A and FIG. 7B.

The beam LBe that passes through the off-state acousto-optical modulation element AM5 and the lens GK7 converges to form a beam waist in the space above the reflecting mirror IM5, and then diverges and enters the lens GK9 coaxially with the optical axis. The front focus of the lens GK9 is arranged to coincide with the position of the rear focus of the lens GK7, and the acousto-optical modulation element AM6 that is switched between the on state (diffracted light generating state) and the off state by the drive signal SSf and is arranged under the Bragg diffraction condition is provided at the position of the rear focus of the lens GK9. The beam LBe that passes through the lens GK9 becomes a parallel light beam with the same beam diameter as when it enters the first stage acousto-optical modulation element AM5.

When the acousto-optical modulation element AM6 is in the on state, as shown in FIG. 20, a beam LB1c (parallel light beam) is generated as a first-order diffraction beam of the beam LBe. The beam LB1c passes through the lens GK10, converges to a beam waist at the position of the incident mirror IM6, is reflected at a right angle by the incident mirror IM6, and enters the lens GK11 in a state coaxial with the optical axis. Here again, the acousto-optical modulation element AM6 is disposed at the position of the front focal point of the lens GK10, and the incident mirror IM6 is disposed at the position of the rear focal point of the lens GK10. Furthermore, the position of the front focal point of the lens GK11 is set to be the same as the position of the rear focal point of the lens GK10 (the position of the incident mirror IM6). Therefore, the beam LB1c that passed through the lens GK11 becomes a parallel light beam again, is reflected by the mirror M59, and enters the beam shape transformation unit 10C as shown in FIG. 4 (FIG. 5A to FIG. 5C) or FIG. 13.

On the other hand, when the first-stage acousto-optical modulation element AM5 is in the on state, a beam LB1b is generated as a first-order diffracted beam of the beam LBe incident on the acousto-optical modulation element AM5. The beam LB1b (parallel light beam) is converged by the lens GK7, reflected by the incident mirror IM5, and incident on the lens GK8 in a state coaxial with the optical axis. Since the front focal point of the lens GK8 is set at the position of the rear focal point of the lens GK7 (the position of the incident mirror IM5), the beam LB1b that passes through the lens GK8 becomes a parallel light beam again, is reflected by the mirror M58, and is incident on the beam shape transformation unit 10B as shown in FIG. 4 (FIG. 5A to FIG. 5C) or FIG. 13.

As described above in FIG. 13, the beam shape transformation units 10B and 10C shown in FIG. 20 may rotate the compression direction of the beams LB1b and LB1c around the optical axis by rotating the beam compression system OM2 by each of the drive mechanisms 20B and 20C. The beams LB1b and LB1c emitted from the beam shape transformation units 10B and 10C are reflected by the mirrors M3B and M3C, respectively, as in FIG. 4, and then pass through the wedge-shaped prisms 12B and 12C to advance so as to form a predetermined angle with the optical axis AXs of the beam switching unit BD1B (BD2B). Also, as in FIG. 4, the beam LB1a from the light source device LS1A is passed between the mirrors M3B and M3C so as to be coaxial with the optical axis AXs of the beam switching unit BD1B (BD2B). The beam LB1a is supplied to the beam switching unit BD1B (BD2B) through the parallel plate 12A in the same manner as in FIG.

With the above configuration, in this modified example, each of the drawing units MU1, MU3 (MU2, MU4) can selectively project one of the circular spot light SPa produced by beam LB1a (LB2a), the slot-shaped (oblong) spot light SPb inclined at -45 degrees produced by beam LB1b (LB2b), and the slot-shaped (oblong) spot light SPb inclined at +45 degrees produced by beam LB1c (LB2c) onto the sheet substrate P, or can simultaneously project both the slot-shaped (oblong) spot light SPb, SPc and the circular spot light SPa onto the sheet substrate P.

When the frequency Fp of the pulse oscillation of each of the light source device LS1A and the light source device LSe is 400 MHz, taking into consideration that the maximum switching frequency Fss of the acousto-optical modulation elements AM5 and AM6 shown in FIG. 20 is about 50 MHz to 100 MHz, the application timing of the drive signals SSc and SSd that switch each of the acousto-optical modulation elements AM5 and AM6 to the on state or off state is executed several pixels earlier than the position of the edge pixel or oblique line pattern to be drawn with the slot-shaped (elliptical) spot light SPb and SPc, for example, as with the drive signals SSa and SSb described in FIG. 15. For example, when each of the spot light SPa, SPb, and SPc is irradiated with n pulses for one pixel in the main scanning direction, the application timing of the drive signals SSc and SSd is controlled to precede by a number of pixels equal to or greater than the value calculated by Fp/n·Fss. Therefore, when Fp=400 MHz, Fss=50 MHz, and n=2, it is sufficient to apply either one of the drive signals (high frequency signals) SSe or SSf at a pixel position preceding by four or more pixels in the main scanning direction.

In this modified example, two acousto-optical modulation elements AM5 and AM6 are used to switch between the spot light SPb and the spot light SPc. For this reason, the control device 100 shown in FIG. 4 is provided with a drive circuit (high-frequency signal application amplifier, etc.) for applying the drive signals SSe and SSf to each of the acousto-optical modulation elements AM5 and AM6, and the drawing data storage unit 100B stores information (bitmap information) for setting the on/off of the drive signals SSe and SSf in association with the pixel map information of the pattern to be drawn. In addition, the beam switching mechanism of the acousto-optical modulation elements AM5 and AM6, lenses GK7 to GK11, and epi-illumination mirrors IM5 and IM6 in this modified example functions as a light splitting unit that splits the beam LBe from one light source device LSe into two beams LB1b and LB1c that travel through different optical paths. Furthermore, mirrors M3B and M3C and wedge-shaped prisms 12B and 12C in this modified example function as a beam combining section that combines two beams LB1b and LB1c whose cross-sectional shapes have been deformed from a circular shape by beam shape modifying sections 10B and 10C.

According to this modification, as shown in Fig. 20, the polarization direction of each of the two beams LB1b (for spot light SPb) and LB1c (for spot light SPc) generated by the beam LBe from one light source device LSe can be aligned with the polarization direction of the beam LB1a (for spot light SPa) generated from one light source device LS1A, so that it is possible to sequentially supply beams from each light source device in a time-division manner to each of the multiple drawing units MU1 to MU4 using the acousto-optical modulation elements AM1 to AM4. Also, in this modification, as in Fig. 17 above, the directionality (long axis direction) of the spot lights SPb and SPc deformed from a circular shape to an anisotropic shape (slot shape, oblong shape) can be changed by the drive mechanisms 20B and 20C, so that it is easy to set the spot shape to suit the angle of the diagonal edge of the pattern to be drawn.

In addition, in this modified example, it is possible to appropriately select and project both the slot-shaped (elliptical) deformed spot light SPb or SPc and the circular spot light SPa, only the deformed spot light SPb or SPc, or only the circular spot light SPa, onto pixels of a pattern edge that extends linearly in the main scanning direction (Yt direction) or the sub-scanning direction (Xt direction), or pixels of a pattern edge that is inclined diagonally with respect to the Yt direction and the Xt direction.

The spot light SPb, SPc described in the first embodiment, the second embodiment, and each modified example above is deformed into a slot shape or an oblong shape with respect to the circular spot light SPa, but it may be deformed into other shapes. For example, the spot light SPb (or SPc) can be made rectangular, square, or rhombus. However, in that case, the minimum spot size is determined from the relationship between the numerical aperture (NA) of the beam B1b projected onto the sheet substrate P via the fθ lens system FT and the second cylindrical lens CYb shown in FIG. 2 and the wavelength λ of the beam B1b (beam from the light source device). Therefore, when the spot light SPb is made rectangular, it is desirable to make the size sufficiently larger than the minimum spot size (for example, three times or more of the minimum spot size) taking into account the effects of diffraction and aberration. Even if the spot light SPb and SPc are both made into the same rectangular shape, if the diagonal directions of the rectangles are different from each other, they are treated as spot lights deformed into different shapes.

Claims (31)

A pattern exposure apparatus including a drawing unit that draws a pattern on a substrate by scanning a spot light of a beam supplied from a light source device in a main scanning direction,
a first light source device that emits a first beam;
a second light source device that emits a second beam;
a beam combining unit that combines the first beam from the first light source device and the second beam from the second light source device so that the first beam and the second beam are incident on the drawing unit;
a beam shape modification unit that changes cross-sectional shapes of the first beam and the second beam incident on the beam combining unit so that a shape of a first spot light by the first beam and a shape of a second spot light by the second beam projected on the substrate are changed from each other;
a control device that controls so that at least an edge portion of a pattern to be drawn on the substrate is drawn with either one or both of the first spot light and the second spot light;
A pattern exposure apparatus comprising: 2. The pattern exposure apparatus according to claim 1,
The beam combining unit combines the first beam and the second beam so that the first beam and the second beam are incident on the drawing unit along an optical axis of the drawing unit. 3. The pattern exposure apparatus according to claim 1,
the first beam from the first light source device and the second beam from the second light source device both have a circular cross-sectional shape;
The beam shape deformation unit includes a first beam shape deformation unit that deforms the cross-sectional shape of the first beam from a circle to a slot shape or an oblong shape, and a second beam shape deformation unit that deforms the cross-sectional shape of the second beam from a circle to a slot shape or an oblong shape. 4. The pattern exposure apparatus according to claim 3,
a direction of a major axis of a cross-sectional shape of the first beam emitted from the first beam shape deformation unit and a direction of a major axis of a cross-sectional shape of the second beam emitted from the second beam shape deformation unit are set to be in different directions from each other so that a major axis direction of the first spot light deformed into the slot shape or the oblong shape or the second spot light deformed into the slot shape or the oblong shape or the second spot light deformed into the slot shape or the oblong shape or the 5. The pattern exposure apparatus according to claim 3,
A pattern exposure apparatus, wherein each of the first beam shape deformation unit and the second beam shape deformation unit includes a beam compression system that compresses a beam having a circular cross-sectional shape in one direction. 6. The pattern exposure apparatus according to claim 5,
The beam compression system of the pattern exposure apparatus includes two cylindrical lenses arranged spaced apart in the optical axis direction so as to convert a beam having a circular cross-sectional shape, which enters as a parallel light beam, into a parallel light beam having a cross-sectional shape deformed into the slot shape or oblong shape and then emit the beam. The pattern exposure apparatus according to any one of claims 3 to 6,
a major axis direction of the first spot light deformed into the slot shape or the oblong shape is set to be inclined in a range of +25 degrees to +65 degrees with respect to the main scanning direction,
a major axis direction of the second spot light deformed into the slot shape or the oblong shape is set to be tilted within a range of −25 degrees to −65 degrees with respect to the main scanning direction. The pattern exposure apparatus according to any one of claims 3 to 7,
a control device for controlling the emission of the first beam from the first light source device and the emission of the second beam from the second light source device so that, when the edge portion of the pattern to be drawn on the substrate is an oblique edge portion extending at an angle with respect to the main scanning direction, one of the first spot light and the second spot light is projected onto the substrate in accordance with the angle of the oblique edge portion. 9. The pattern exposure apparatus according to claim 8,
a third light source device that emits a third beam having a circular cross-sectional shape;
The beam combining unit causes the third beam to be incident along the optical axis of the drawing unit, together with the first beam and the second beam which have been deformed into the slot shape or the oblong shape, so that a circular third spot light by the third beam is projected on the substrate. 10. The pattern exposure apparatus according to claim 9,
The control device includes:
a pattern exposure apparatus that controls the emission of the first beam from the first light source device and the emission of the second beam from the second light source device so that, for the diagonal edge portions of a pattern to be drawn on the substrate, either the first spot light or the second spot light is projected onto the substrate, and controls the emission of the third beam from the third light source device so that, for pattern portions other than the diagonal edge portions, the emission of the third beam from the third light source device is projected onto the substrate. A pattern exposure apparatus including a drawing unit that draws a pattern on a substrate by scanning a spot light of a beam supplied from a light source device in a main scanning direction,
a light splitting unit that splits a beam having a circular cross section emitted from the light source device into a first beam and a second beam;
a first beam shape deformation unit provided in an optical path of the first beam and configured to deform a cross-sectional shape of the first beam from a circular shape to form a first spot light shape on the substrate by projecting the first beam into a first shape;
a second beam shape deformation unit provided in an optical path of the second beam and configured to deform a cross-sectional shape of the second beam from a circular shape to form a shape of a second spot light on the substrate by projecting the second beam into a second shape different from the first shape;
a beam combining unit that combines the first beam from the first beam shape transforming unit and the second beam from the second beam shape transforming unit so as to be incident on the drawing unit;
a control device that controls the pattern to be drawn on the substrate to be drawn using either the first spot light or the second spot light;
A pattern exposure apparatus comprising: 12. The pattern exposure apparatus according to claim 11,
The beam combining unit combines the first beam and the second beam so that the first beam and the second beam are incident on the drawing unit along an optical axis of the drawing unit. 13. The pattern exposure apparatus according to claim 11,
the first beam shape deformation unit deforms a cross-sectional shape of the first beam such that the first shape of the first spot light is deformed into a slot shape or an oblong shape having a major axis inclined with respect to the main scanning direction,
the second beam shape deformation unit deforms a cross-sectional shape of the second beam so that the second shape of the second spot light is deformed into a slot shape or an oblong shape having a major axis inclined opposite to the first shape with respect to the main scanning direction. 14. The pattern exposure apparatus according to claim 13,
A pattern exposure apparatus, wherein each of the first beam shape deformation unit and the second beam shape deformation unit includes a beam compression system that compresses a beam having a circular cross-sectional shape in one direction to make the cross-sectional shape into a slot shape or an oblong shape. 15. The pattern exposure apparatus according to claim 14,
The beam compression system of the pattern exposure apparatus includes two cylindrical lenses arranged apart in the optical axis direction so as to convert a beam having a circular cross-sectional shape, which enters as a parallel light beam, into a parallel light beam having a cross-sectional shape deformed into the slot shape or oblong shape and output the beam. The pattern exposure apparatus according to any one of claims 13 to 15,
a major axis direction of the first spot light deformed into the slot shape or the oblong shape is set to be inclined in a range of +25 degrees to +65 degrees with respect to the main scanning direction,
a major axis direction of the second spot light deformed into the slot shape or the oblong shape is set to be tilted within a range of −25 degrees to −65 degrees with respect to the main scanning direction. The pattern exposure apparatus according to any one of claims 13 to 16,
the beam synthesis unit is composed of a polarizing beam splitter having transmissivity or reflectivity according to the direction of linearly polarized light,
a first beam shape modification unit that modifies a direction of linear polarization of the first beam from the first beam shape modification unit and a second beam from the second beam shape modification unit in a complementary manner; 18. The pattern exposure apparatus according to claim 17,
The control device controls the electro-optical element so that, when an edge portion of a pattern to be drawn on the substrate is an oblique edge portion extending at an angle with respect to the main scanning direction, one of the first spot light and the second spot light is projected onto the substrate in accordance with the angle of the oblique edge portion. 20. The pattern exposure apparatus according to claim 18,
a second light source device that emits a third beam having a circular cross-sectional shape;
The beam combining unit causes the third beam to be incident along the optical axis of the drawing unit, together with the first beam or the second beam that has been deformed into the slot shape or the oblong shape, so that a circular third spot light by the third beam is projected onto the substrate. 20. The pattern exposure apparatus according to claim 19,
a control device that controls the emission of the beam from the light source device and the driving of the electro-optical element so that, for the diagonal edge portions of a pattern to be drawn on the substrate, either the first spot light or the second spot light is projected onto the substrate, and that controls the emission of the third beam from the second light source device so that, for pattern portions other than the diagonal edge portions, the third spot light is projected onto the substrate. A pattern exposure method using a drawing unit that draws a pattern on a substrate by scanning a spot light of a beam supplied from a light source device in a main scanning direction in accordance with pixel information for each pixel defined in drawing data, comprising:
a pattern exposure method, wherein when an edge pixel that becomes an edge portion of a pattern extending obliquely intersecting the main scanning direction is included in the row of pixels scanned in the main scanning direction, a shape of the spot light projected onto at least the edge pixel is set to a slot shape or an oblong shape whose major axis is tilted in a direction along the oblique extension direction of the pattern. The pattern exposure method according to claim 21,
A pattern exposure method, wherein a shape of the spot light projected onto the pixels other than the edge pixels is set to be circular. The pattern exposure method according to claim 22,
A pattern exposure method, wherein an effective size of the circular spot light projected onto the substrate is set within a range of ±50% of a size of the pixel defined on the substrate. The pattern exposure method according to claim 23,
A pattern exposure method, wherein an effective dimension in a major axis direction of the slot-shaped or oblong spot light projected onto the substrate is set to be equal to a diagonal dimension of the pixel defined on the substrate. The patterned exposure method according to any one of claims 22 to 24,
the light source device includes a first light source device that emits a first beam that becomes the slot-shaped or oblong first spot light whose major axis direction is inclined in a range of +25 degrees to +65 degrees with respect to the main scanning direction, and a second light source device that emits a second beam that becomes the slot-shaped or oblong second spot light whose major axis direction is inclined in a range of -25 degrees to -65 degrees with respect to the main scanning direction,
A pattern exposure method in which, when drawing an edge pixel corresponding to the edge portion, either the first beam from the first light source device or the second beam from the second light source device is controlled to be supplied to the drawing unit. The patterned exposure method according to any one of claims 22 to 24,
the light source device includes a first light source device that emits a beam for generating the slot-shaped or oblong-shaped first spot light whose major axis direction is inclined within a range of +25 degrees to +65 degrees with respect to the main scanning direction and the slot-shaped or oblong-shaped second spot light whose major axis direction is inclined within a range of -25 degrees to -65 degrees with respect to the main scanning direction, and a second light source device that emits a beam for generating the circular third spot light. 27. The pattern exposure method according to claim 26,
a first beam for generating the first spot light and a second beam for generating the second spot light, a cross-sectional shape of each of the first beam and the second beam is deformed into the slot shape or the oblong shape, and then one of the first beam and the second beam is supplied to the drawing unit so as to be aligned with an optical axis within the drawing unit. The pattern exposure method according to claim 27,
A pattern exposure method, comprising: splitting the beam from the first light source device into the first beam and the second beam using an electro-optical element that switches the polarization direction of the beam from the first light source device by electrical control; and a polarizing beam splitter that splits the beam that has passed through the electro-optical element into a transmitting optical path and a reflecting optical path depending on the polarization state. The pattern exposure method according to claim 27,
a first acousto-optical modulation element and a second acousto-optical modulation element are provided, the first acousto-optical modulation element and the second acousto-optical modulation element are arranged so that the beam from the first light source device passes through them in series;
A pattern exposure method, comprising: using a first-order diffracted beam of the beam generated when only the first acousto-optical modulation element is in an on-state as the first beam; and using a first-order diffracted beam of the beam generated when only the second acousto-optical modulation element is in an on-state as the second beam. A pattern exposure apparatus including a drawing unit that draws a pattern on a substrate by scanning a spot light of a beam supplied from a light source device in a main scanning direction,
a first light source device that emits a first beam;
a second light source device that emits a second beam;
a first beam shape deformation unit that receives the first beam, deforms a cross-sectional shape of the first beam, and guides the first beam to the drawing unit ;
a second beam shape deformation unit that receives the second beam and deforms a cross-sectional shape of the second beam to guide the second beam to the drawing unit ,
The first beam shape deformation unit deforms a cross-sectional shape of the first beam from a circular shape to a non-circular shape,
A pattern exposure apparatus, wherein a first spot light projected onto the substrate corresponding to the deformed first beam and a second spot light projected onto the substrate corresponding to the deformed second beam have mutually different shapes . 31. The pattern exposure apparatus according to claim 30,
The first beam shape deformation unit deforms a cross-sectional shape of the first beam from a circular shape to a slot shape or an oblong shape.

Images