WO2021206044A1

WO2021206044A1 - Pattern-forming device and pattern-forming method

Info

Publication number: WO2021206044A1
Application number: PCT/JP2021/014467
Authority: WO
Inventors: 加藤正紀
Original assignee: 株式会社ニコン
Priority date: 2020-04-06
Filing date: 2021-04-05
Publication date: 2021-10-14
Also published as: CN115380253A; KR20220150942A; TW202303300A; TWI781572B; JPWO2021206044A1; JP7435748B2; TW202205031A

Abstract

A pattern-forming device for forming a pattern in a predetermined region on a substrate moving in a first direction, the device comprising: a first alignment system that optically detects first substrate marks formed on the substrate and spaced at predetermined intervals in the first direction, in a first detection region set upstream of the predetermined region in the first direction; a second alignment system that optically detects, in a second detection region set upstream of the predetermined region in the first direction, second substrate marks which are formed on the substrate, are spaced at predetermined intervals in the first direction, and are separated from the first substrate marks by a predetermined distance in a second direction orthogonal to the first direction; a reference-indicating member extending in the second direction along the first and second alignment systems, and having reference-indicating marks formed at portions which correspond to the first and second detection regions, respectively, in the second direction; and optical combination members that are provided in the first and second alignment systems, respectively, and that perform combination of light beams so as to cause light beams from the reference-indicating marks to pass through an optical path through which light beams from the substrate marks pass.

Description

Pattern forming device and pattern forming method

The present invention relates to a pattern forming apparatus for transferring a pattern onto a substrate by a transfer unit for forming a pattern in a forming region extending in the width direction of the substrate, and a pattern forming method.

In order to transfer a pattern onto a long flexible sheet substrate, a plurality of head assemblies (transfer units) are arranged in a short direction (width direction) orthogonal to the long direction of the sheet substrate, and the sheet substrate is moved in the long direction. A drawing device that draws (exposes) a pattern with a plurality of head assemblies while moving the pattern is disclosed in, for example, the following Japanese Patent Application Laid-Open No. 2006-0987626.

In the drawing apparatus of JP-A-2006-09876, as shown in FIGS. 3 and 7 to 9, for example, four camera units 52 are placed on the upstream side of the exposure position for alignment. The positions of the four camera units 52 are calibrated by the calibration scale 42 provided in the scanning transport unit 26 with the detection unit, which is arranged in the width direction of the substrate) 28 and before the exposure operation. In Japanese Patent Application Laid-Open No. 2006-0987626, a long base material 28 is adsorbed and supported in a plane at an exposure position, and an endless belt 33 that can move in the long length direction along the surface of the flat base material 28 is provided. .. The endless belt 33 is provided on a scanning transport unit 26 that is moved in a long direction by a linear moving mechanism 20 and a moving table 21. Further, the scanning transport unit 26 is provided with a nip roller pair 30, a nip drive roller pair 32, an output guide roller 40, and the like, and each of the four camera units 52 is provided with the base material 28 supported on the endless belt 33. The position can be calibrated. In the drawing apparatus of JP-A-2006-09876, each of the four camera units 52 is configured to be movable in the width direction of the base material 28 corresponding to each position of the mark M formed on the base material 28. , The calibration scale 42 is used to correctly set the moving position.

In the apparatus configuration of JP-A-2006-09876, the calibration operation (so-called calibration) can be performed in a state where the scanning transport unit 26 is moved and the calibration scale 42 is positioned directly under the four camera units 52. However, in order to eliminate the need to remove the base material 28 to be exposed from the apparatus, the endless belt 33, the nip roller pair 30, the nip drive roller pair 32, the output guide roller 40, etc. mounted on the scanning transport unit 26, etc. The accompanying mechanism is required, the weight of the scanning transport unit 26 is increased, and the linear moving mechanism 20 or the like that controls the movement of the scanning transport unit 26 needs to be made stable with high rigidity, which leads to an increase in the size of the apparatus. Further, since the calibration scale 42 is attached to the scanning transport unit 26 that moves one-dimensionally by the linear movement mechanism 20, the positioning of the scanning transport unit 26 is reproduced in accordance with the alignment accuracy (position detection accuracy of the mark M). It is also necessary to make the accuracy such as sex sufficiently small.

A first aspect of the present invention is a pattern forming apparatus that forms a pattern for an electronic device in a predetermined region on a substrate that is supported by a moving mechanism and moves in a first direction, and is supported by the moving mechanism. A pattern forming mechanism that forms the pattern on the surface of the substrate within a pattern forming region in which the dimension of the second direction orthogonal to the first direction is set longer than the dimension of the first direction on the surface of the substrate. And, the first substrate marks formed on the surface of the substrate at predetermined intervals along the first direction are set on the upstream side of the pattern forming region with respect to the moving direction of the substrate. The first alignment system, which is optically detected inside, and the first alignment system are formed on the surface of the substrate at predetermined intervals along the first direction at a predetermined distance from the first substrate mark in the second direction. A second alignment system that optically detects the second substrate mark within the second detection region set on the upstream side of the pattern formation region with respect to the moving direction of the substrate, and the first alignment system. And the second alignment system are extended in the second direction by the low expansion material, and correspond to the portion corresponding to the first detection region and the second detection region in the second direction. From the index pattern, a reference index member in which an index pattern is formed in each of the portions, and an optical path provided in each of the first alignment system and the second alignment system and through which light from the substrate mark passes. It is provided with a synthetic optical member that synthesizes light so as to pass through.

A second aspect of the present invention is a pattern forming method in which a new pattern is superimposed on a base pattern for an electronic device formed in a predetermined region on a substrate moving in the first direction. With respect to the movement of the substrate in the first direction, the first alignment system set on the upstream side of the pattern forming region by the pattern forming mechanism for forming the new pattern in the predetermined region of the substrate. A first mark detection step of sequentially optically detecting a plurality of first substrate marks formed on the substrate at predetermined intervals along the first direction within one detection region, and upstream of the pattern formation region. Within the second detection region of the second alignment system, which is on the side and is set at a predetermined interval in the second direction orthogonal to the first direction from the first detection region, the first detection region is placed on the substrate. A second mark detection step of sequentially optically detecting a plurality of second substrate marks formed at predetermined intervals along the direction, and the second alignment system along the first alignment system and the second alignment system. One of the two reference index marks formed at each of the positions corresponding to the design distance between the first detection region and the second detection region on the reference index member extending in the direction. The first reference index mark is simultaneously detected with the first substrate mark via a synthetic optical member arranged in the optical path of the first alignment system during the first mark detection step, and the two The other second reference index mark of the reference index mark is simultaneously detected with the second substrate mark via the synthetic optical member arranged in the optical path of the second alignment system during the second mark detection step. By doing so, a first measurement step of measuring each position of the first substrate mark and the second substrate mark with reference to the reference index mark, and within the predetermined region on the substrate by the pattern forming mechanism. When forming the new pattern, the first adjustment for adjusting the position of the new pattern based on the positions of the first substrate mark and the second substrate mark measured in the first measurement step. Including the process.

It is a perspective view which shows the schematic whole structure of the pattern drawing apparatus EX by 1st Embodiment. It is a figure which showed concretely the arrangement of the drawing units U1 to U6 and the arrangement of an alignment system ALGn in the pattern drawing apparatus EX of FIG. It is a perspective view which shows the detailed structure in the drawing unit U1 as a representative among the drawing units U1 to U6 in the pattern drawing apparatus EX of FIG. It is a perspective view which shows the arrangement relation of the rotary drum DR, the alignment system ALGn, and the reference bar member RB shown in FIG. FIG. 4 is a view showing the arrangement relationship of the objective lens system OBL, the plane mirror Mb, the beam splitter BS1, and the reference bar member RB of the alignment system ALGn shown in FIG. 4 in a plane parallel to the XY plane of FIG. 2 is a perspective view showing an overall schematic configuration of the alignment systems ALGn (ALG1 to ALG4) shown in FIGS. 2 to 5. It is a figure which showed the arrangement relation of the drawing lines SL1 to SL6 and the detection area AD1 to AD4 shown in FIG. 4, and the arrangement of the encoder measurement system which measures the change of the rotation angle of a rotating drum DR. FIG. 8 shows an example of the arrangement relationship and support structure of the drawing units U1 to U6, the scale disk SDa, the encoder heads EHa1, EHa2, EHa3, and the reference bar member RB, and FIG. 8A is a view of the periphery of the scale disk SDa. FIG. 8B is a partial cross-sectional view of the end face when the structure of FIG. 8A is broken along the central plane CPo, as viewed from the + X direction side to the −X direction side. FIG. 9 is a diagram showing a detailed configuration of the alignment system ALGn (ALG1 to ALG4), and FIG. 9A shows a configuration of an optical system in which the plane mirror Mb is omitted from the configuration shown in FIG. 9B shows an example of the illumination field diaphragm FAn provided in the illumination system ILU, and FIG. 9C shows an enlarged image RNn'of the reference mark RNn and an enlarged image MKn'of the alignment mark MKn imaged by the image sensor DIS. It is a figure which shows an example. FIG. 10A is a diagram showing an example of the arrangement of the reference marks RM1 to RM4 formed at four positions of the reference bar member RB, and FIG. 10B exaggerates an example of the arrangement relationship between the imaging region DIS'and the reference mark RM1. FIG. 10C is a diagram showing an exaggerated example of the arrangement relationship between the imaging region DIS'and the reference mark RM2. It is a block diagram which shows the schematic structure of a part of the control apparatus provided in the pattern drawing apparatus EX by this embodiment. It is a flowchart which shows an example of the series flow of the operation sequence of the pattern drawing apparatus EX by this embodiment. It is a figure which shows the structure of the alignment system ALGn by 1st Embodiment. It is a figure which exaggerated the installation error ΔCn of the alignment system ALGn based on the reference mark RMn on the reference bar member RB shown in FIG. 10A. It is a figure which shows the arrangement example of the detection area ADn of the reference pattern FMa, FMb, FMc ..., and the alignment system ALGn formed on the outer peripheral surface DRs of the rotary drum DR developed in a plane. It is a figure showing an example of arrangement of each image pickup area DIS'(detection area ADn) of 6 drawing lines SLn, 4 alignment system ALGn, and reference pattern FMa. It is a figure which shows how the spot light SP which generates the drawing line SL1 relatively two-dimensionally scans the area containing the linear pattern Fxc1 of the reference pattern FMa on a rotating drum DR. It is a figure explaining how to determine the arrangement error relative to the adjacent drawing line SL2 with the drawing line SL1 in which the spot light SP is scanned as a reference at the time of drawing. FIG. 5 is a diagram schematically exaggerating calibration information (arrangement error, etc.) determined or set by step 304 in the flowchart of FIG. 12. It is a figure which shows the structure by the modification 1 when the alignment system ALGn of the drawing apparatus shown in FIG. 7 is increased from 4 lines. FIG. 5 is a diagram illustrating a modification 2 in the case where the arrangement relationship of the drawing lines SL1 to SL6 shown in FIGS. 7, 16 and 20 on the substrate P is modified to provide an overlap region at the joint portion. be. An example of the state of the same pattern (pixel array on the two-dimensional drawing data) to be overexposed in the overlap region OL12 between the scanning start point side of the drawing line SL1 and the scanning start point side of the drawing line SL2 is shown. It is a figure. It is a figure which shows the optical composition of the alignment system ALGn by 2nd Embodiment. It is a figure which shows the optical composition of the alignment system ALGn by 3rd Embodiment. It is a figure which shows the optical arrangement by the modification 3 when the configuration of the beam splitter BS1 (compositing optical member) in the alignment system ALGn shown in FIG. 24 is modified. FIG. 26 is a diagram showing the configuration of the reference bar member RB according to the fourth embodiment, FIG. 26A shows the configuration of the reference bar member RB on the reference surface RBa, and FIG. 26B is the reference in FIG. 26A. The CC-CC arrow cross section of the bar member RB is shown, and FIG. 26C shows an example of the configuration of the reference mark RMn formed on the reference surface RBa. It is a figure which shows the schematic structure of the pattern drawing apparatus which used the digital mirror device (DMD) as the maskless exposure apparatus according to the 5th Embodiment. It is a figure which shows the arrangement example in the XY plane of the projection area IAn by each drawing unit Un in the pattern drawing apparatus shown in FIG. 27, and the detection area ADn by each of the alignment system ALGn. FIG. 29 is a diagram illustrating a modification 5 relating to the configuration of the alignment system ALGn, FIG. 29A shows an optical configuration modified based on the alignment system ALGn shown in FIG. 24, and FIG. 29B is a beam splitter. It is a graph which shows an example of the wavelength selection characteristic of BS1. It is a figure which shows the structure by the modification 6 which uses the light from two solid-state light sources which have different wavelength characteristics as the illumination system ILU which supplies the illumination light to the alignment system ALGn of FIG. 29A. It is a figure which shows the structure by the modification 7 which deformed the structure of the beam splitter BS1 of the alignment system ALGn, and the arrangement direction of a reference bar member RB. FIG. 32 shows the difference in the deformed state depending on the support structure of the reference bar member RB, and FIG. 32A shows the structure in which the vicinity of both ends of the reference bar member RB in the longitudinal direction is supported from below by the beam FJ1 which makes line contact with each other. 32C shows a structure in which the vicinity of one end of the reference bar member RB in the longitudinal direction is supported from below by a beam FJ1 that makes line contact, and the other end is fastened (fixed) to the device frame FJ2. FIG. 32C shows the reference. A structure is shown in which both ends of the bar member RB in the longitudinal direction are fastened (fixed) to the device frame FJ2. It is a partial perspective view which shows the modification of the support structure of the reference bar member RB by the

support members

103A and 103B described with FIG.

The substrate processing apparatus (pattern forming apparatus) and the substrate processing method (pattern forming method) according to the aspect of the present invention will be described in detail below with reference to the attached drawings, with reference to suitable embodiments. It should be noted that the aspects of the present invention are not limited to these embodiments, but include those with various changes or improvements. That is, the components described below include those that can be easily assumed by those skilled in the art and those that are substantially the same, and the components described below can be appropriately combined. In addition, various omissions, substitutions or changes of components can be made without departing from the gist of the present invention.

[First Embodiment]
FIG. 1 is a perspective view showing a schematic configuration of a pattern forming apparatus (pattern drawing apparatus) EX that transfers a pattern to a substrate (irradiated body) P as a substrate processing apparatus according to the first embodiment. , International Publication No. 2017/191777, International Publication No. 2018/061633. In the following description, unless otherwise specified, the XYZ Cartesian coordinate system with the gravity direction as the Z direction is set, and the X direction, the Y direction, and the Z direction are set according to the arrows shown in the figure.

The pattern drawing device EX is used in a device manufacturing system that manufactures an electronic device by exposing a fine pattern for an electronic device to a photosensitive functional layer such as a photoresist coated on the substrate P. In device manufacturing, for example, in order to manufacture electronic devices such as flexible displays, film-like touch panels, film-like color filters for liquid crystal display panels, flexible wiring, or flexible sensors, in addition to pattern drawing devices, Multiple types of manufacturing equipment are used. In the device manufacturing system, the substrate P is delivered from a supply roll (not shown) in which a flexible sheet-shaped substrate (sheet substrate) P is wound in a roll shape, and various processes are continuously performed on the delivered substrate P. It has a so-called roll-to-roll (Roll To Roll) production method in which the substrate P after various treatments is wound up with a recovery roll (not shown). Therefore, at least on the substrate P during the manufacturing process, a large number of patterns corresponding to the unit device (one display panel, etc.) to be the final product are arranged in a state of being connected with a predetermined gap in the transport direction of the substrate P. Will be done. The substrate P has a strip shape in which the elongated direction is the moving direction (transporting direction) of the substrate P and the short direction orthogonal to the elongated direction is the width direction of the substrate P.

For the substrate P, for example, a resin film, a foil made of a metal or alloy such as stainless steel, or the like is used. Examples of the material of the resin film include polyethylene resin, polypropylene resin, polyester resin, ethylene vinyl copolymer resin, polyvinyl chloride resin, cellulose resin, polyamide resin, polyimide resin, polycarbonate resin, polystyrene resin, and vinyl acetate resin. Of these, those containing at least one or more may be used. Further, the thickness and rigidity (Young's modulus) of the substrate P are within a range in which the substrate P does not have creases or irreversible wrinkles due to buckling when passing through the transport path of the device manufacturing system or the pattern drawing device EX. It should be. As the base material of the substrate P, films such as PET (polyethylene terephthalate), PEN (polyethylene naphthalate), and polyimide having a thickness of about 25 μm to 200 μm are typical of suitable sheet substrates. The substrate P is a single layer of ultrathin glass having a thickness of about 30 to 100 μm manufactured by a float method or the like, a laminate obtained by laminating the above resin film, metal foil, or the like on the ultrathin glass, or A piece of paper containing nanocellulose and having its surface smoothed may be used.

The photosensitive functional layer applied to the surface of the substrate P is applied as a solution on the substrate P and dried to form a layer (film). A typical photosensitive functional layer is a photoresist (liquid or dry film), but as a material that does not require development treatment, the photosensitive functional layer is modified in terms of the liquid-repellent property of the portion irradiated with ultraviolet rays. There are silane coupling agents (SAM), photosensitive reducing agents in which the plating reducing group is exposed on the portion irradiated with ultraviolet rays, and the like. When a photosensitive silane coupling agent is used as the photosensitive functional layer, the pattern portion exposed to ultraviolet rays on the substrate P is modified from liquid-repellent to liquid-friendly. Therefore, a thin film transistor (TFT) or the like can be obtained by selectively coating a conductive ink (an ink containing conductive nanoparticles such as silver or copper) or a liquid containing a semiconductor material on the liquid-forming portion. It is possible to form a pattern layer that serves as an electrode, a semiconductor, an insulation, or a wiring for connection.

When a photosensitive reducing agent is used as the photosensitive functional layer, the plating reducing group is exposed on the pattern portion (or the unexposed pattern portion) exposed to ultraviolet rays on the substrate P. Therefore, a pattern layer made of palladium is formed (precipitated) by electroless plating in which the substrate P is immediately immersed in a plating solution containing palladium ions or the like for a certain period of time after exposure. Such a plating process is an additive (addition type) process, but in addition, an etching process as a subtractive (subtraction type) process may be premised. In that case, the substrate P sent to the pattern drawing apparatus EX uses PET or PEN as the base material, and a metallic thin film such as aluminum (Al) or copper (Cu) is vapor-deposited on the entire surface or selectively of the substrate P. It is assumed that a photoresist layer is laminated on top.

The pattern drawing device EX shown in FIG. 1 is a direct drawing type exposure device that does not use a mask, that is, a so-called spot scanning type exposure device, and the substrate P conveyed from the process device in the previous process is used as a process device in the subsequent process. It is conveyed in the elongated direction at a predetermined speed toward (including a single processing unit or a plurality of processing units). In synchronization with the transfer, the pattern drawing device EX has one of the signal line and power supply line wiring pattern constituting the electronic device, the electrode constituting the TFT, the semiconductor region, the through hole, and the like on the photosensitive functional layer of the substrate P. An optical pattern corresponding to the pattern shape of the above is formed by high-speed scanning (main scanning) of spot light whose intensity is modulated according to drawing data in the Y direction and movement of the substrate P in the long direction (secondary scanning). ..

In FIG. 1, the pattern drawing device EX has a rotary drum DR that supports the substrate P for subscanning and conveys it in a long direction, and a pattern for each portion of the substrate P that is supported by the rotary drum DR in a cylindrical surface shape. A plurality of (here, 6) drawing units Un (U1 to U6) for exposure are provided, and each of the plurality of drawing units Un (U1 to U6) is a pulsed beam LB (pulse beam) for exposure. While scanning the spot light one-dimensionally (main scanning) with the polygon mirror PM (scanning member) in the Y direction on the irradiated surface (photosensitive surface) of the substrate P, the intensity of the spot light is increased at high speed according to the drawing data. Modulate (on / off). As a result, an optical pattern corresponding to a predetermined pattern such as an electronic device, a circuit, or a wiring is drawn and exposed on the irradiated surface of the substrate P. Further, since the substrate P is continuously conveyed along the elongated direction, the exposed region (device forming region) on the substrate P on which the pattern is exposed by the pattern drawing device EX is in the elongated direction of the substrate P. A plurality of can be set at a predetermined interval (margin) along the line. In the present embodiment, the pattern forming mechanism is configured by each or all of the six drawing units U1 to U6.

As shown in FIG. 1, the rotating drum DR has a central axis AXo extending in the Y direction and extending in a direction intersecting the direction in which gravity acts, and a cylindrical outer peripheral surface having a constant radius from the central axis AXo. The rotary drum DR rotates around the central axis AXo while supporting (holding close contact) a part of the substrate P by bending it in a cylindrical surface shape in the elongated direction following its outer peripheral surface (circumferential surface). The substrate P is conveyed in the long direction. The rotating drum DR supports a region (part) on the substrate P on which the beam LB (spot light) from each of the plurality of drawing units Un (U1 to U6) is projected by its outer peripheral surface. On both sides of the rotating drum DR in the Y direction, shafts (not shown) supported by bearings so as to rotate the rotating drum DR around the central axis AXo are provided. Rotational torque from a rotational drive source (for example, a motor, a reduction mechanism, etc.) (not shown) is applied to the shaft, and the rotary drum DR rotates at a constant rotational speed around the central axis AXo.

The light source device (pulse light source device) LS generates and emits a pulsed beam (pulse beam, pulse light, laser) LB. This beam LB has sensitivity to the photosensitive functional layer of the substrate P and has a peak wavelength (for example, any wavelength of 405 nm, 365 nm, 355 nm, 344 nm, 308 nm, 248 nm, etc.) in the wavelength band of 410 to 200 nm. It is ultraviolet light. The light source device LS emits a pulsed beam LB at an FPL of any frequency (oscillation frequency, predetermined frequency) in the range of 100 MHz to 400 MHz, for example, under the control of a drawing control device (not shown). In the present embodiment, the light source device LS is a laser light source device that generates ultraviolet light by a wavelength conversion element. Specifically, a semiconductor laser element that generates pulsed light in the infrared wavelength region, a fiber amplifier, and a wavelength conversion element (harmonic) that converts the amplified pulsed light in the infrared wavelength region into pulsed light in the ultraviolet wavelength region. A fiber amplifier laser light source composed of a generating element) or the like. By configuring the light source device LS in this way, it is possible to obtain high-intensity pulsed light of ultraviolet rays having a light emission time of one pulse light of a dozen picoseconds to a few tens of picoseconds or less. The light source device LS is used as a fiber amplifier laser light source, and the pulse generation of the beam LB is turned on / off at high speed according to the state of the pixel bits (logical value "0" or "1") constituting the drawing data (spot light). The (intensity-modulated) configuration is disclosed in International Publication No. 2015/166910 and International Publication No. 2017/057415. It is assumed that the beam LB emitted from the light source device LS has a thin parallel luminous flux having a beam diameter of about 1 mm or about half of the beam diameter.

The beam LB emitted from the light source device LS includes a selection optical element OSn (OS1 to OS6) as a plurality of switching elements, a plurality of reflection mirrors M1 to M12, and a plurality of epi-illumination mirrors (also referred to as selection mirrors) Imn (also referred to as a selection mirror). It is selectively (alternatively) supplied to each of the drawing units Un (U1 to U6) via the beam switching unit composed of the IM1 to IM6) and the absorber TR or the like. The selection optical elements OSn (OS1 to OS6) have transparency with respect to the beam LB, and are driven by an ultrasonic signal to efficiently generate only one of the ± 1st-order diffracted light of the incident beam LB. It is composed of an acousto-optic modulator (AOM) arranged under black diffraction conditions. The plurality of selection optical elements OSn and the plurality of epi-illumination mirrors Imn are provided corresponding to each of the plurality of drawing units Un. For example, the selection optical elements OS1 and the epi-illumination mirror IM1 are provided corresponding to the drawing unit U1, and similarly, the selection optical elements OS2 to OS6 and the epi-illumination mirrors IM2 to IM6 correspond to the drawing units U2 to U6, respectively. It is provided.

The optical path of the beam LB from the light source device LS is bent in a zigzag shape in a plane parallel to the XY plane by the reflection mirrors M1 to M12, and is guided to the absorber TR. Hereinafter, the case where the selection optical elements OSn (OS1 to OS6) are all off (a state in which an ultrasonic signal is not applied and primary diffracted light is not generated) will be described in detail. Although not shown in FIG. 1, a relay system using a plurality of lenses is provided in the beam optical path from the reflection mirror M1 to the absorber TR. The relay system is specifically arranged between the selection optical elements OS1 to OS6 arranged in series along the optical path of the beam LB from the light source device LS, as disclosed in International Publication No. 2017/057415. , Each of the six selection optical elements OS1 to OS6 is optically coupled to each other (imaging relationship). Further, each relay system maintains the diameter of the beam LB as a thin parallel luminous flux of 1 mm to 0.5 mm at each position of the six selection optical elements OS1 to OS6, and the diameter is 0 at the intermediate position in each relay system. Converge so that the beam waist is 2 mm or less. Each of the epi-illumination mirrors IM1 to IM6 is arranged at the position of the beam waist in the optical path of each relay system.

In FIG. 1, the beam LB from the light source device LS travels in the −X direction, is reflected by the reflection mirror M1 in the −Y direction, and is incident on the reflection mirror M2. The beam LB reflected in the + X direction by the reflection mirror M2 passes straight through the selection optical element OS5 and reaches the reflection mirror M3. The beam LB reflected in the −Y direction by the reflection mirror M3 is reflected in the −X direction by the reflection mirror M4, passes straight through the selection optical element OS6, and reaches the reflection mirror M5. The beam LB reflected in the −Y direction by the reflection mirror M5 is reflected in the + X direction by the reflection mirror M6, passes straight through the selection optical element OS3, and reaches the reflection mirror M7. The beam LB reflected in the −Y direction by the reflection mirror M7 is reflected in the −X direction by the reflection mirror M8, and then passes straight through the selection optical element OS4 to reach the reflection mirror M9. The beam LB reflected in the −Y direction by the reflection mirror M9 is reflected in the + X direction by the reflection mirror M10, and then passes straight through the selection optical element OS1 to reach the reflection mirror M11. The beam LB reflected in the −Y direction by the reflection mirror M11 is reflected in the −X direction by the reflection mirror M12, and then passes straight through the selection optical element OS2 and is guided to the absorber TR. This absorber TR is an optical trap that absorbs the beam LB in order to suppress leakage of the beam LB to the outside, and is provided with a temperature control (air cooling or water cooling) mechanism so as to reduce heat generation due to absorption of light energy. ing.

When an ultrasonic signal (high frequency signal) is applied, each selection optical element OSn emits first-order diffracted light obtained by diffracting an incident beam LB (0th-order light) at a diffraction angle corresponding to a high-frequency frequency. It is generated as (Beam LBn). Therefore, the beam emitted as the primary diffracted light from the selection optical element OS1 becomes LB1, and similarly, the beam emitted as the primary diffracted light from each of the selection optical elements OS2 to OS6 becomes LB2 to LB6. As described above, each selection optical element OSn (OS1 to OS6) functions to deflect the optical path of the beam LB from the light source device LS. Further, in the present embodiment, only one of the selection optical elements OSn (OS1 to OS6) is turned on for a certain period of time (a state in which a high frequency signal is applied), which is not shown. It is controlled by the drawing control device. When one selected optical element OSn for selection is in the ON state, about 10 to 20% of 0th-order light that travels straight without being diffracted by the optical element OSn for selection remains, but it is finally determined by the absorber TR. Be absorbed.

Each of the selection optical elements OSn is installed so as to deflect the beam LBn (LB1 to LB6), which is the deflected primary diffracted light, in the −Z direction with respect to the traveling direction of the incident beam LB. The beam LBn (LB1 to LB6) deflected and emitted by each of the selection optical elements OSn is an epi-illumination mirror Imn (position of the beam waist) provided at a position (beam waist position) separated from each of the selection optical elements OSn by a predetermined distance. It is projected on IM1 to IM6). Each epi-illumination mirror Imn reflects the incident beam LBn (LB1 to LB6) in the −Z direction to guide the beam LBn (LB1 to LB6) to the corresponding drawing units Un (U1 to U6).

The configuration, function, operation, etc. of each selection optical element OSn are the same as each other, and each of the plurality of selection optical element OSn is incident by turning on / off the drive signal (high frequency signal) from the drawing control device. A switching (beam selection) operation is performed to turn on / off the generation of diffracted light (beams LB1 to LB6) obtained by diffracting the beam LB. By such a switching operation of each selection optical element OSn, the beam LB from the light source device LS can be guided to any one drawing unit Un, and the drawing unit Un on which the beam LBn is incident can be switched. .. In this way, a configuration in which a plurality of selection optical elements OSn are arranged in series (serially) in the optical path of the beam LB and the beam LBn is supplied to the corresponding drawing unit Un in a time-division manner is described in International Publication No. 2015/166910. It is disclosed in.

The order in which each of the selection optical elements OSn (OS1 to OS6) constituting the beam switching unit is turned on for a certain period of time is the scanning start timing by the spot light set in each of the drawing units Un (U1 to U6). Determined by order. In the present embodiment, any one of the drawing units U1 to U6 is synchronized by synchronizing the rotation speeds of the polygon mirrors PM provided in each of the six drawing units U1 to U6 and the phases of the rotation angles. One reflective surface of the polygon mirror in the above can be time-division-switched so as to perform one spot scan on the substrate P. Therefore, as long as the phases of the rotation angles of the polygon mirrors of the drawing unit Un are synchronized in a predetermined relationship, the spot scanning order of the drawing unit Un may be any order. In the configuration of FIG. 1, three drawing units U1, U3, and U5 are arranged side by side in the Y direction on the upstream side of the substrate P in the transport direction (the direction in which the outer peripheral surface of the rotating drum DR moves in the circumferential direction), and the substrate P is arranged. Three drawing units U2, U4, and U6 are arranged side by side in the Y direction on the downstream side in the transport direction.

In this case, the pattern drawing for one exposed area on the substrate P is started from the odd-numbered drawing units U1, U3, and U5 on the upstream side, and when the substrate P is sent for a certain length, the even-numbered drawing on the downstream side is drawn. Since the units U2, U4, and U6 also start pattern drawing, the spot scanning order of the drawing unit Un can be set to U1 → U3 → U5 → U2 → U4 → U6 → U1 → ... .. Therefore, the order in which each of the selection optical elements OSn (OS1 to OS6) is turned on for a certain period of time is also determined in the order of OS1 → OS3 → OS5 → OS2 → OS4 → OS6 → OS1 → ...

As shown in FIG. 1, each of the drawing units U1 to U6 is provided with a polygon mirror PM for main scanning the incident beams LB1 to LB6. In the present embodiment, each of the polygon mirror PMs of each drawing unit Un is synchronously controlled so as to maintain a constant rotation angle phase with each other while precisely rotating at the same rotation speed. Thereby, the timings of the main scans of the beams LB1 to LB6 projected from the drawing units U1 to U6 on the substrate P (the main scan period of the spot light) can be set so as not to overlap each other. Therefore, by controlling the on / off switching of each of the selection optical elements OSn (OS1 to OS6) provided in the beam switching unit in synchronization with the rotation angle position of each of the six polygon mirror PMs, the light source device Efficient exposure processing can be performed by distributing the beam LB from the LS to each of the plurality of drawing units Un in a time-division manner. The synchronization control between the phase matching of each rotation angle of the six polygon mirror PMs and the on / off switching timing of each of the selection optical elements OSn (OS1 to OS6) is also disclosed in International Publication No. 2015/166910. ing.

As shown in FIG. 1, the pattern drawing apparatus EX is a so-called multi-head type direct drawing exposure method in which a plurality of drawing units Un (U1 to U6) having the same configuration are arranged. Each of the drawing units Un draws a pattern for each partial region defined in the Y direction (main scanning direction) of the substrate P supported by the outer peripheral surface (circumferential surface) of the rotating drum DR. Each drawing unit Un (U1 to U6) focuses (converges) the beam LBn on the substrate P while projecting the beam LBn from the beam switching unit onto the substrate P (on the irradiated surface of the substrate P). As a result, the beams LBn (LB1 to LB6) projected on the substrate P become spot light having a diameter of 2 to 4 μm. Further, by rotating the polygon mirror PM of each drawing unit Un, each spot light of the beams LBn (LB1 to LB6) projected on the substrate P is scanned in the main scanning direction (Y direction). By scanning the spot light, a linear drawing line (scanning line) SLn (n = 1, 2, ..., 6) for drawing a pattern for one line is defined on the substrate P. NS. The drawing line SLn is a scanning locus of the spot light of the beam LBn on the substrate P.

Each drawing line SLn (SL1 to SL6) of the plurality of drawing units Un (U1 to U6) sandwiches a central surface parallel to the YZ surface including the central axis AXo of the rotating drum DR, and is 2 in the circumferential direction of the rotating drum DR. Arranged in a staggered arrangement in a row, the odd-numbered drawing lines SL1, SL3, and SL5 are located on the irradiated surface of the substrate P on the upstream side (-X direction side) of the substrate P in the transport direction with respect to the central surface. Moreover, they are arranged in a row at a predetermined interval along the Y direction. The even-numbered drawing lines SL2, SL4, and SL6 are located on the irradiated surface of the substrate P on the downstream side (+ X direction side) of the substrate P in the transport direction with respect to the central surface, and are predetermined along the Y direction. They are arranged in a row at intervals of. Therefore, the odd-numbered drawing units U1, U3, and U5 and the even-numbered drawing units U2, U4, and U6 are provided symmetrically with respect to the central plane when viewed in the XZ plane (when viewed from the Y direction). There is.

Regarding the X direction (conveying direction of the substrate P), the odd-numbered drawing lines SL1, SL3, SL5 and the even-numbered drawing lines SL2, SL4, SL6 are separated from each other, but in the Y direction (width direction of the substrate P). , Main scanning direction), the patterns drawn on the substrate P are set so as to be spliced together without being separated from each other. The drawing lines SL1 to SL6 are set so as to be substantially parallel to the width direction of the substrate P, that is, the central axis AXo of the rotating drum DR. Note that joining the drawing lines SLn in the Y direction means that the positions of the ends of the drawing lines SLn in the Y direction are adjacent to each other or partially overlapped. When the ends of the drawing lines SLn are overlapped in the Y direction, for example, the length of each drawing line SLn is 1 to 5% in the Y direction including the drawing start point or the drawing end point. It is good to duplicate.

The plurality of drawing units Un (U1 to U6) share a scanning area (main scanning range section) in the Y direction so as to cover the width direction dimension of the exposure region (pattern formation region) on the substrate P in total. doing. For example, assuming that the main scanning range (the length of the drawing line SLn) in the Y direction by one drawing unit Un is about 30 to 60 mm, drawing is possible by arranging the six drawing units U1 to U6 in the Y direction. The width of the exposed area in the Y direction is widened to about 180 to 360 mm. In principle, the length (drawing range length) of each drawing line SLn (SL1 to SL6) is the same. That is, in principle, the scanning distances of the spot light SPs of the beams LBn scanned along each of the drawing lines SL1 to SL6 are the same.

Each drawing unit Un (U1 to U6) includes a telecentric fθ lens system (scanning optical system for drawing) FT that injects a beam LBn that is reflected by each reflecting surface RP of the polygon mirror PM and deflected in the main scanning direction. , The beam LBn emitted from the fθ lens system FT and projected onto the substrate P is set to travel toward the central axis AXo of the rotating drum DR when viewed in the XZ plane. As a result, the main ray of the beam LBn traveling from each drawing unit Un (U1 to U6) toward the substrate P is directed to the tangent plane at the position of the drawing line SLn on the curved surface of the substrate P in the XZ plane. It is projected toward the substrate P so that it is always vertical. That is, the beams LBn (LB1 to LB6) projected on the substrate P are scanned in a telecentric state with respect to the main scanning direction and the sub-scanning direction (circumferential direction along the outer peripheral surface of the rotating drum DR) of the spot light SP. ..

FIG. 2 shows the arrangement of the rotary drum DR of the pattern drawing apparatus EX and the six drawing units U1 to U6 shown in FIG. 1, the alignment mark formed on the substrate P, the reference pattern formed on the surface of the rotary drum DR, and the like. It is a figure which concretely showed the arrangement of a plurality of alignment systems ALGn (n is an integer of 2 or more) which detects, and the setting of the Cartesian coordinate system XYZ in FIG. 2 is the same as FIG. The basic arrangement of the rotating drum DR, the drawing units U1 to U6, and the alignment system ALGn shown in FIG. 2 is disclosed in, for example, International Publication No. 2016/152758 and International Publication No. 2017/199658.

Shafts Sft supported by an annular bearing so as to rotate around the central axis AXo are provided on both sides of the rotating drum DR that supports the substrate P in an angle range of about 180 degrees in the Y direction, and the shaft Sft is not provided. It is joined to the rotation shaft of the rotation drive source shown in the figure. Further, the plane including the central axis AXo and parallel to the YZ plane is defined as the central plane CPo. When viewed from the Y direction (width direction of the substrate P), the odd-numbered drawing units U1, U3, and U5 and the even-numbered drawing units U2, U4, and U6 are symmetrically arranged with the central surface CPo in between. As shown in FIG. 2, in the plane parallel to the XZ plane of the Cartesian coordinate system XYZ, the drawing unit U1 (and U3, U5) is tilted counterclockwise by a certain angle θc from the central plane CPo, and the drawing unit U2 (and U3, U5) And U4, U6) are tilted clockwise from the central surface CPo by a certain angle θc. Since the configurations of the drawing units U1 to U6 are the same, the configuration of the drawing unit U1 is shown in FIG. 3 as a representative.

In FIG. 3, the beam LB1 (parallel light flux having a diameter of about 1 mm to 0.5 mm whose intensity is modulated according to the drawing data) supplied from the epi-illumination mirror IM1 shown in FIG. 1 is finally spotlighted on the substrate P. FIG. 5 is a perspective view showing a detailed configuration of a drawing unit U1 including an fθ lens system FT that collects light as an SP and a polygon mirror PM that mainly scans a spot light SP in the Y direction to form a drawing line SL1. It is disclosed in No. 2019/082850. The optical axis AXf1 passing through the fθ lens system FT from the polygon mirror PM of the drawing unit U1 (and U2 to U6) is tilted in the Cartesian coordinate system XYZ as shown in FIG. , The Cartesian coordinate system XtYtZt tilted with respect to the Cartesian coordinate system XYZ is set. In the Cartesian coordinate system XtYtZt, the Yz direction is the same as the Y direction, and the Zt direction is the traveling direction of the main ray (center ray) of the beam LB1 incident on the drawing unit U1 from the epi-illumination mirror IM1, or the position of the drawing line SL1. The Xt direction is the direction of the optical axis AXf1 passing through the fθ lens system FT. The optical axis of each fθ lens system FT of the even-numbered drawing units U2, U4, and U6 is the optical axis AXf2.

In FIG. 3, in the drawing unit U1 (and U2 to U6), a mirror M30, a lens L6, a lens L7, a tiltable parallel flat plate HVP made of quartz, lenses L8, L9, a mirror M31, a polarizing beam splitter PBS, and an opening. Aperture AP, 1/4 wavelength plate QW, mirror M32, first cylindrical lens CYa, lens L10, mirror M33, lens L11, mirror M34, M35, M36, 8-sided polygon mirror PM, fθ lens system FT, mirror M37, The second cylindrical lens CYb is arranged in that order. The mirror M30 reflects the beam LB1 at 90 degrees so that the traveling direction of the incident beam LB1 is in the −Xt direction. The lenses L6, L7, L8, and L9 arranged along the optical path of the beam LB1 reflected by the mirror M30 are a few mm or more of the thin beam LB1 (diameter is about 1 mm to 0.5 mm) reflected by the mirror M30. It constitutes a beam expander system that expands to a parallel light flux with a diameter of (range of 5 to 10 mm).

The parallel flat plate HVP is provided in the optical path between the lenses L6 to L9 of the beam expander system, and is configured to be rotatable (tilted) around the rotation axis AXh parallel to the Zt axis. By changing the amount of inclination of the parallel flat plate HVP, the position of the spot light SP projected on the substrate P can be effectively changed to the sub-scanning direction (Xt direction, the sub-scanning direction which is the moving direction of the substrate P). It can be shifted in a distance range of several times to ten and several times the diameter φp. The polarizing beam splitter PBS incident on the beam LB1 (parallel luminous flux) which is magnified through the lens L9 and reflected by the mirror M31 in the −Yt direction. Assuming that the beam LB1 is linearly P-polarized, the polarization beam splitter PBS reflects the beam LB1 at the polarization separation surface with an intensity of 99% or more and directs the beam LB1 to the aperture stop AP in the subsequent stage. The beam LB transmitted through the circular aperture of the aperture stop AP is converted from linearly polarized light to circularly polarized light when transmitted through the quarter wave plate QW.

The beam LB1 (parallel luminous flux) transmitted through the 1/4 wave plate QW is reflected by the mirror M32 in the −Zt direction, is incident on the first cylindrical lens CYa (the bus is parallel to the Yt axis), and is formed on the surface Pv in space. The width in the Xt direction is extremely small, and the light is focused on a slit-shaped intensity distribution extending in the Yt direction with a length of several mm (same as the opening diameter of the aperture throttle AP). The beam LB1 converged only in the one-dimensional direction by the surface Pv passes through the spherical lens L10 of the first group of the two-disc spherical lens system, is reflected by the mirror M33 in the + Xt direction, and then is reflected in the + Xt direction, and then the two-disc spherical lens system. It advances in the + Xt direction through the spherical lens L11 in the rear group. The beam LB1 after being emitted from the spherical lens L11 is reflected in the + Zt direction by the mirror M34 and then reflected in the + Yt direction by the mirror M35. The mirror M34 and the mirror M35 are arranged so that the main ray (center ray) of the beam LB1 traveling in the + Yt direction from the mirror M35 and the optical axis AXf1 of the fθ lens system FT are orthogonal to each other in a plane parallel to the XtYt plane. There is. The beam LB1 traveling in the + Yt direction from the mirror M35 is reflected by the mirror M36 arranged on the opposite side of the mirror M35 with the optical axis AXf1 of the fθ lens system FT interposed therebetween, and is projected onto the reflection surface RPa of the polygon mirror PM.

Due to the action of the first cylindrical lens CYa and the two-disc spherical lens system, the beam LB1 incident on the mirror M34 immediately after passing through the spherical lens L11 becomes a state of almost parallel light beam in the Zt direction and converges in the Yt direction. It becomes a state of light beam. In FIG. 3, the spherical lens system is composed of two spherical lenses L10 and L11 for adjusting the distance between the principal points, but it may be composed of only one spherical lens.

The reflective surface of the mirror M36 is arranged at a narrowing angle of 22.5 ° with respect to the surface parallel to the Zt axis and parallel to the XtZt surface and including the optical axis AXf1. As a result, the main ray (center ray) of the beam LB1 directed from the mirror M36 toward the reflecting surface RPa of the polygon mirror PM, that is, the extension of the optical axis of the first cylindrical lens CYa and the spherical lens system (lenses L10 and L11), is formed on the mirror M36. The optical axis from to the polygon mirror PM is set at an angle of 45 ° with respect to the optical axis AXf1 of the fθ lens system FT in a plane parallel to the XtYt plane. Further, in FIG. 3, the beam LB1 reflected by the mirror M36 and directed toward the reflecting surface RPa of the polygon mirror PM is in a convergent light beam state so as to be focused on the reflecting surface RPa of the polygon mirror PM in the Zt direction, and is in a state of convergent light beam XtYt. In the plane parallel to the plane, the light beam is almost parallel, and on the reflecting surface RPa, the intensity distribution extends in a slit shape in the main scanning direction, that is, in the tangential direction of the inscribed circle centered on the rotation center axis AXp of the polygon mirror PM. It is condensed so as to be.

The beam LB1 reflected by the reflecting surface RPa of the polygon mirror PM passes through the telecentric fθ lens system FT, and then is reflected by the mirror M37 at a right angle in the −Zt direction to the second cylindrical lens CYb (the direction of the bus is Yt). It is incident on the substrate P and is focused as spot light SP on the substrate P. In the present embodiment, the optical axis AXf1 of the fθ lens system FT, which is bent at a right angle in the −Zt direction by the mirror M37 and becomes perpendicular to the surface of the substrate P (the outer peripheral surface of the rotating drum DR), and toward the mirror M30. The central ray of the beam LB1 incident in the −Zt direction is set to be coaxial with the line segment LE1 parallel to the Zt axis (the line segments LE2 to LE6 are used for each of the other drawing units U2 to U6). ing. With such a setting, when the drawing line SL1 is tilted by a small amount in the substrate P (a plane parallel to the XtYt plane), each optical member from the mirror M30 to the second cylindrical lens CYb shown in FIG. 3 is integrally formed. The entire supporting housing (unit support frame) can be slightly rotated around the line segment LE1. As described above, for a mechanism that enables the entire support frame of the drawing unit U1 (the same applies to the other units U2 to U6) to rotate slightly around the line segment LE1 (LE2 to LE6), for example, International Publication No. 2016/152758 It is disclosed in.

Further, in the present embodiment, the intensity of the reflected light generated when the spot light SP is projected on the surface of the irradiated object (the substrate P or the outer peripheral surface of the rotating drum DR) installed on the surface to be scanned is detected. Therefore, a photoelectric sensor DTR and a lens system GF are provided. The reflected light (particularly normal reflected light) from the surface of the irradiated body is the second cylindrical lens CYb, the fθ lens system FT, the reflecting surface RPa of the polygon mirror PM, the mirrors M36, M35, M34, the spherical lens L11, and the mirror M33. It returns to the polarizing beam splitter PBS via the spherical lens L10, the first cylindrical lens CYa, the mirror M32, the 1/4 wavelength plate QW, and the aperture aperture AP. The spot light SP projected on the surface of the irradiated body is circularly polarized light, and the reflected light also contains a large amount of circularly polarized light components. Therefore, the reflected light passes through the 1/4 wavelength plate QW and becomes the polarized beam splitter PBS. When heading, its polarization characteristics are converted to linear S-polarized light. Therefore, the reflected light from the surface of the irradiated body passes through the polarization separation surface of the polarization beam splitter PBS and is incident on the lens system GF. The light receiving surface of the photoelectric sensor DTR is set to have an optically conjugate relationship with the spot light SP on the scanning surface so that the reflected light from the irradiated body is focused on the light receiving surface of the photoelectric sensor DTR by the lens system GF. Will be done.

Although not shown in FIG. 3, as disclosed in International Publication No. 2015/166910 or International Publication No. 2016/152758, the reflective surface of the polygon mirror PM on which the drawing beam LB1 is projected. Sending for the origin sensor to output a pulsed origin signal indicating that each reflective surface of the polygon mirror PM is at the angular position immediately before the start of drawing to the reflective surface RPb immediately before the rotation direction of RPa. A light beam is projected. Further, the detailed internal configuration of the drawing unit U1 shown in FIG. 3 is the same for the other drawing units U2 to U6, but each of the even-numbered drawing units U2, U4, and U6 has the drawing unit U1 of FIG. Is installed in a direction rotated by 180 degrees around the line segment LE1.

Here, the configuration of the pattern drawing device EX will be further described with reference to FIG. 2 again. In the odd-numbered drawing units U3 and U5 including the drawing unit U1 shown in FIG. 3, the extension lines of the line segments LE1, LE3, and LE5 (that is, the extension lines of the optical axis AXf1 of the fθ lens system FT) are shown in FIG. The line segments LE1, LE3, and LE5 are installed so as to be inclined counterclockwise by an angle −θc with respect to the central surface CPo while facing the rotation central axis AXo of the rotating drum DR when viewed from the Y direction of 2. On the other hand, in the even-numbered drawing units U2, U4, and U6, the extension lines of the line segments LE2, LE4, and LE6 (that is, the extension lines of the optical axis AXf2 of the fθ lens system FT) are viewed from the Y direction of FIG. The line segments LE2, LE4, and LE6 are installed so as to be tilted clockwise by an angle + θc with respect to the central surface CPo while moving toward the rotation central axis AXo of the rotating drum DR. The angle ± θc is set to be as small as possible within a range in which the odd-numbered drawing units U1, U3, and U5 and the even-numbered drawing units U2, U4, and U6 do not spatially interfere with each other (do not collide).

In the present embodiment, a plurality of alignment systems ALGn are arranged in the Y direction at predetermined intervals, and each includes an objective lens system for detecting a mark or the like on the substrate P. The detection region (observation field of view) set on the substrate P via the objective lens system is drawn by each of the drawing units U1 to U6 with respect to the moving direction of the substrate P (circumferential direction of the outer peripheral surface of the rotating drum DR). It is arranged on the upstream side of the positions of lines SL1 to SL6. The extension lines of the respective optical axes AXs of the objective lens system passing through the center of the detection region (observation field of view) are directed toward the rotation center axis AXo of the rotating drum DR, and the surface of the substrate P is located at the position of the detection area (observation field of view). Alternatively, it is set to be perpendicular to the outer peripheral surface of the rotating drum DR. A reference bar member RB as a reference index member forming a reference mark (reference index mark) is attached near the tip of the alignment system ALGn. The reference mark of the reference bar member RB calibrates the mutual positional relationship of the detection areas (observation fields of view) by each of the objective lens systems, or the mutual positional relationship of the drawing lines SL1 to SL6 by each of the drawing units U1 to U6. It is also used when measuring the interval (baseline length) and positional relationship in the circumferential direction (moving direction of the substrate P) between the positions of the drawing lines SL1 to SL6 and each position of the plurality of detection areas. When viewed in a plane parallel to the XZ plane, each optical axis AXs of the alignment system ALGn has an angle θa larger than the angle θc of the drawing lines SL1, SL3, SL5 by the odd-numbered drawing units U1, U3, and U5, respectively. Only the central surface CPo is set to tilt counterclockwise.

FIG. 4 is a perspective view showing the arrangement relationship between the rotating drum DR, the alignment system ALGn, and the reference bar member RB shown in FIG. 2, and the Cartesian coordinate system XYZ is the same as the Cartesian coordinate system XYZ of FIG. Set to the same. In the present embodiment, the four alignment systems ALG1 to ALG4 are linearly arranged in the Y direction at predetermined intervals, and the optical axes AXs of the objective lens system OBL of the alignment system ALG1 are the objective lens system OBL and the substrate P (rotation). It is bent by a planar mirror Mb arranged between the outer peripheral surface of the drum DR and the cube-shaped beam splitter BS1 and is set to pass through the center point of the detection region (observation field) AD1 on the substrate P. Although the reference numerals are omitted in FIG. 4, each of the other alignment systems ALG2, ALG3, and ALG4 is provided with the same objective lens system OBL, plane mirror Mb, and cube-type beam splitter BS1 (synthetic optical member) for alignment. The optical axes AXs of the systems ALG2, ALG3, and ALG4 are also set to pass through the center points of the detection regions (observation fields of view) AD2, AD3, and AD4 set on the substrate P, respectively. In the present embodiment, the alignment system ALG1 is used as the first alignment system and the alignment system ALG4 is used as the second alignment system, but any one of the four alignment systems ALG1 to ALG4 is used as the first alignment system. The alignment system may be used, and any one of the remaining three alignment systems may be used as the second alignment system.

The reference bar member RB is elongated in the Y direction from a material having a low coefficient of thermal expansion (Invar, ceramics, quartz, etc.) and is attached in the vicinity of each beam splitter BS1 of the four alignment systems ALG1 to ALG4. As the material of the reference bar member RB, it is desirable to use ceramics that can also reduce the weight, and in particular, it is composed of three components of _{magnesium oxide (MgO), aluminum oxide (Al 2} O ₃ ), and silicon dioxide (SiO _2). It is recommended to use aluminum-based ceramics. A detection region AR1 is set at a position corresponding to the detection region AD1 on the substrate P of the alignment system ALG1 on the reference surface RBa facing the beam splitter BS1 of the reference bar member RB. A reference mark (reference pattern) observable by the objective lens system OBL is formed in the detection region AR1 of the reference surface RBa via the beam splitter BS1 and the plane mirror Mb. Therefore, the alignment system ALGn in the present embodiment is the alignment mark (or the outer peripheral surface of the rotating drum DR) on the substrate P that appears in the detection region AD1 via the beam splitter BS1 arranged on the tip side of the objective lens system OBL. The reference pattern formed above) and the reference mark on the reference surface RBa of the reference bar member RB set in the detection area AR1 can be observed simultaneously or selectively.

FIG. 5 shows the arrangement relationship of the objective lens system OBL of the alignment system ALGn, the plane mirror Mb, the cubic cube-shaped beam splitter BS1, and the reference bar member RB in a plane parallel to the XY plane of FIG. It is a figure. The Cartesian coordinate system XYZ is set to be the same as the Cartesian coordinate system XYZ in FIG. 4 or FIG. The optical axis AXs extended from the objective lens system OBL is bent obliquely downward (-Z direction) by the plane mirror Mb. When viewed in the XZ plane, the reflective surface of the plane mirror Mb is tilted counterclockwise by an angle θk with respect to the plane perpendicular to the optical axis AXs extending from the objective lens system OBL. In the case of the present embodiment, the optical axis AXs passing through the objective lens system OBL is constant in the XZ plane because the alignment system ALGn is arranged in the space below the odd-numbered drawing units U1, U3, and U5 as shown in FIG. It is tilted with respect to the XY plane by the angle of. As described with reference to FIG. 2, since the odd-numbered drawing units U1, U3, and U5 line segments LE1, LE3, and LE5 are tilted by an angle −θc with respect to the central surface CPo, the objective lens system is adjusted accordingly. The optical axes AXs passing through the OBL are also tilted by an angle −θc with respect to the XY plane. Therefore, when viewed in the XZ plane, the reflecting surface of the plane mirror Mb is tilted counterclockwise by an angle − (θk + θc) with respect to the plane parallel to the YZ plane.

The optical axis AXs that are folded back by the plane mirror Mb and directed toward the beam splitter BS1 are tilted by an angle of 2θk in the −Z direction with respect to the original optical axis AXs, with respect to a plane parallel to the XY plane of the Cartesian coordinate system XYZ. Therefore, it is tilted by an angle − (θc + 2θk). The beam splitter BS1 has an optical splitting surface Bsp in which the slopes of two right-angled prisms (for example, made of quartz) are joined to each other, and the cross-sectional shape seen in the XZ surface is formed into a substantially square shape as a whole. The surface PBa on the plane mirror Mb side of the beam splitter BS1 and the surface PBc on the reference bar member RB side are parallel to each other and are molded at 45 degrees with respect to the optical division surface Bsp. Further, the surface PBb on the side of the beam splitter BS1 facing the substrate P (rotating drum DR) is orthogonal to each of the surface PBa and the surface PBc, and is molded at 45 degrees with respect to the optical splitting surface Bsp. The surface PBb of the beam splitter BS1 is set parallel to the surface (tangible plane) of the substrate P, and the surface PBc is set parallel to the reference surface RBa of the reference bar member RB.

Further, the optical axis AXs that vertically passes through the surface PBa of the beam splitter BS1 is set so as to vertically pass through the surface PBc and be perpendicular to the reference surface RBa of the reference bar member RB. Further, the optical axis AXs that passes vertically through the plane PBa of the beam splitter BS1 and is reflected at 90 degrees by the optical splitting plane Bsp so as to pass vertically through the plane PBb and be perpendicular to the surface (tangent plane) of the substrate P. Set. The optical path length from the tip surface of the objective lens system OBL to the surface of the substrate P is set to be equal to the optical path length from the tip surface of the objective lens system OBL to the reference surface RBa of the reference bar member RB, and the beam splitter BS1 is the objective. It is arranged so as not to block the optical path between the lens system OBL and the plane mirror Mb. As is clear from FIG. 5, the working distance (working distance) from the tip of the objective lens system OBL to the surface of the substrate P or the reference surface RBa is set long because the plane mirror Mb and the beam splitter BS1 are interposed. .. The working distance is set to 10 cm or more as an example.

Next, the overall schematic configuration of the alignment system ALGn (ALG1 to ALG4) will be described with reference to the perspective view of FIG. The Cartesian coordinate system XYZ of FIG. 6 is set to be the same as the Cartesian coordinate system XYZ of FIGS. 4 and 5. The alignment system ALGn is provided with a lens system Gb, a beam splitter BS2, an imaging unit IMS, and an illumination system ILU that are arranged coaxially with the optical axis AXs of the objective lens system OBL. The illumination light ILb from the illumination system ILU is reflected by the mirror Ma, enters from the lower surface (-Z direction) of the beam splitter BS2, is reflected by the optical division surface, passes through the lens system Gb, and passes through the lens system Gb to pass through the objective lens system OBL. Incident in. The illumination light ILb passes through the objective lens system OBL in the detection areas ADn (AD1 to AD4) on the substrate P side and the detection areas ARn (AR1 to AR4) on the reference surface RBa side of the reference bar member RB with a uniform illuminance distribution. Illuminate. The wavelength range of the illumination light ILb is set to a long wavelength side outside the photosensitive wavelength range of the photosensitive functional layer (photoresist or the like) formed on the substrate P, and is set to, for example, 470 nm to 650 nm.

When the alignment mark (board mark) MKn (MK1 to MK4) on the board P appears in the detection area ADn (AD1 to AD4) with the movement of the board P, the reflected light from the detection area ADn including the mark MKn appears. The two-dimensional intensity distribution of is changed. The reflected light from the detection region ADn passes through the beam splitter BS1, the planar mirror Mb, the objective lens system OBL, the lens system Gb, and the beam splitter BS2 in this order, and is received by the imaging unit IMS. The image pickup unit IMS has a two-dimensional image pickup element such as a CCD or CMOS, and images an image of the mark MKn. Further, since the detection region ARn on the reference surface RBa of the reference bar member RB is also illuminated by the illumination light ILb, the reflected light from the detection region ARn on the reference surface RBa is also the beam splitter BS1, the plane mirror Mb, and the objective lens. The light is received by the imaging unit IMS through the system OBL, the lens system Gb, and the beam splitter BS2 in this order.

The alignment mark MK1 is formed on one end side of the width direction (Y direction) of the substrate P at regular intervals (for example, 5 to 20 mm) along the long direction of the substrate P, and the alignment mark MK4 is formed on the substrate P. It is formed on the other end side in the width direction (Y direction) of P at regular intervals (for example, 5 to 20 mm) along the long direction of the substrate P. The alignment marks MK2 and MK3 are inside the width direction of the substrate P, and a blank portion between a plurality of exposed areas arranged at regular intervals in the long direction or a pattern for a device is formed in the exposed area. It is formed in a blank area that is not.

When the size of the rectangular detection region ADn set on the substrate P and the size of the rectangular detection region ARn set on the reference surface RBa of the reference bar member RB are aligned, the image sensor of the image pickup unit IMS The image of the alignment marks MKn (MK1 to MK4) on the substrate P and the image of the reference marks RMn (RM1 to RM4) formed in the detection region ARn on the reference surface RBa are simultaneously formed in the image pickup surface of the above. Since there is an image timing, the arrangement and shape of the alignment mark MKn and the reference mark RNn are set so that the image of the alignment mark MKn and the image of the reference mark RNn do not overlap each other in the imaging surface. Details of the arrangement and shape of the alignment marks MKn (MK1 to MK4) and the reference marks RMn (RM1 to RM4) will be described later.

FIG. 7 shows the arrangement relationship between the drawing lines SL1 to SL6 by each of the drawing units U1 to U6 shown in FIG. 4 and the detection areas AD1 to AD4 of the alignment systems ALG1 to ALG4, and the change in the rotation angle of the rotating drum DR. It is a figure which looked at the arrangement of the encoder measurement system which measures (the position change in the circumferential direction of a substrate P) in the plane parallel to the XY plane of the Cartesian coordinate system XYZ. Scale disks SDa and SDb (same diameter) are fixed to each of the shafts Sft at both ends of the rotating drum DR in the Y direction so as to rotate together with the rotating drum DR coaxially with the central axis AXo. The diameters of the scale disks SDa and SDb are preferably the same as the diameter of the rotating drum DR, but the relative difference in diameter may be within ± 20%. Diffraction grating-like scales Gm are formed on the cylindrical outer peripheral surfaces of the scale disks SDa and SDb at a constant pitch in the circumferential direction. The scale Gm may be formed directly on the outer peripheral surfaces of the rotary drum DR on both end sides in the Y direction.

Around the scale disk SDa, three optical encoder heads EHa1, EHa2, and EHa3 for measuring the amount of movement of the scale Gm in the circumferential direction are provided side by side in the circumferential direction of the outer peripheral surface of the scale disk SDa, and the scale disk SDb Three optical encoder heads EHb1, EHb2, and EHb3 for measuring the amount of movement of the scale Gm in the circumferential direction are provided side by side in the circumferential direction of the outer peripheral surface of the scale disk SDb. The reading position of the scale Gm by the pair of encoder heads EHa1 and EHb1 in the circumferential direction is set to be the same as the angular position in the circumferential direction of the detection areas AD1 to AD4 arranged in a row in the Y direction of each of the alignment systems ALG1 to ALG4. Will be done. Similarly, the reading position of the scale Gm by the pair of encoder heads EHa2 and EHb2 in the circumferential direction is set to be the same as the angular position in the circumferential direction of the odd-numbered drawing lines SL1, SL3, and SL5 arranged in a row in the Y direction. The reading position of the scale Gm by the pair of encoder heads EHa3 and EHb3 in the circumferential direction is set to be the same as the angular position in the circumferential direction of the even-numbered drawing lines SL2, SL4, and SL6 arranged in a row in the Y direction. NS. An encoder measurement system having such an arrangement of encoder heads is disclosed in, for example, International Publication No. 2013/146184, but it is possible to minimize the measurement error.

Further, in FIG. 7, assuming that the maximum dimension in the Y direction that can be continuously exposed by the six drawing lines SL1 to SL6 is Way and the dimension (short length) in the width direction (Y direction) of the substrate P is LPy, the short length of the substrate P The length LPy is smaller than the dimension of the outer peripheral surface of the rotating drum DR in the Y direction, and is smaller than the dimension of the distance between the detection regions AD1 and AD4 of the alignment systems ALG1 and ALG4 set on both ends in the Y direction in the Y direction. Set to be large. Alignment marks MK1 formed at regular intervals (for example, 5 to 20 mm) in the X direction (sub-scanning direction) at the end in the −Y direction on the substrate P are positioned so as to appear in the detection region AD1 of the alignment system ALG1. Alignment marks MK4 formed in the + Y direction on the substrate P at regular intervals (for example, 5 to 20 mm) in the X direction (sub-scanning direction) appear in the detection region AD4 of the alignment system ALG4. It is formed in such a position. Further, the spacing dimension of the detection regions AD1 and AD4 of the alignment systems ALG1 and ALG4 in the Y direction is set within the range of the maximum dimension WAY. In the present embodiment, the linear region formed by each of the drawing lines SL1 to SL6 or the rectangular region surrounded by the entire drawing lines SL1 to SL6 correspond to the pattern forming region.

FIG. 8 shows an example of the arrangement relationship and support structure of the drawing units U1 to U6, the scale disk SDa, the encoder heads EHa1, EHa2, EHa3, and the reference bar member RB, and FIG. 8A shows an example of the orthogonal coordinate system XYZ as in FIG. It is a view of the circumference of the scale disk SDa from the −Y direction side to the + Y direction side, and FIG. 8B shows the end face when the structure of FIG. 8A is broken along the central plane CPo in FIG. It is a partial cross-sectional view seen from the direction side toward the −X direction side.

In FIGS. 8A and 8B, the odd-numbered drawing units U1, U3, and U5 and the even-numbered drawing units U2, U4, and U6 are attached to the support frame portion 100 so as to face each other with the central surface CPo in between. The support frame portion 100 is formed in a rod shape extending in the Y direction in parallel with the central axis AXo of the rotary drum DR, and is fixed to the main body frame of the pattern drawing device EX. The support frame portion 100 pivotally supports each of the drawing units U1 to U6 around the respective line segments LE1 to LE6 in a minute rotatable manner. Such a structure is disclosed in International Publication No. 2016/152758.

On the end side (scale disk SDa side) of the support frame portion 100 on the −Y direction side, an arc-shaped support plate portion 103A to which the three encoder heads EHa1, EHa2, EHa3 and the reference bar member RB are fixed, and A support plate portion 102A for fixing the support plate portion 103A to the end portion side of the support frame portion 100 on the −Y direction side is provided. Although not shown, the encoder heads EHb1, EHb2, EHb3 and the reference bar member RB are similarly fixed to the end side (scale disk SDb side) of the support frame portion 100 on the + Y direction side in an arc shape. A support plate portion 103B and a support plate portion 102B for fixing the support plate portion 103B to the end portion on the + Y direction side of the support frame portion 100 are provided. Further, in order to connect the two arc-shaped

support plate portions

103A and 103B located parallel to each other in the Y direction, the connecting

bar members

104a, 104b and 104c extending in the Y direction are provided at three locations along the arc. It is provided in each of. In the above configuration, the support frame portion 100, the

support plate portions

102A, 102B, 103A, 103B, and the connecting

bar members

104a, 104b, 104c are integrated by using a metal material or a ceramic material having a low coefficient of thermal expansion. It is assembled as a metrology frame (measurement frame) with high mechanical rigidity and extremely small structural deformation due to temperature changes.

The reference bar member RB is bridged to each of the two arc-shaped

support plate portions

103A and 103B separated in the Y direction in the posture shown in FIGS. 4 to 6 above in the Cartesian coordinate system XYZ. Is fixed to. The fine adjustment mechanism 106 provided in each of the

support plate portions

103A and 103B finely adjusts the posture such as the two-dimensional position of the reference bar member RB in the XZ plane and the inclination in the XZ plane on the order of several microns or less. adjust. By individually adjusting the fine adjustment mechanism 106 on the support plate 103A side and the fine adjustment mechanism 106 on the support plate 103B side, the inclination of the reference surface RB of the reference bar member RB can be adjusted in the YZ plane and the reference surface RBa. It is also possible to adjust the parallelism with the central axis AXo of. The fine adjustment mechanism 106 is mainly used during calibration and maintenance work of the device.

Next, the details of the optical system of the alignment system ALGn (ALG1 to ALG4) will be described with reference to FIG. The alignment system ALGn shown in FIG. 9A has the same configuration as that shown in FIG. 6, but for the sake of simplicity, the optical path ahead of the objective lens system OBL is omitted by omitting the plane mirror Mb in FIG. It has been expanded. The illumination system ILU of the alignment system ALGn includes the detection regions ADn (AD1 to AD4) on the substrate P (or the outer peripheral surface of the rotating drum DR) defined by the objective lens system OBL and the reference surface RBa of the reference bar member RB. A plurality of interchangeable lenses are set in an optically conjugate relationship (imaging relationship) with each of the detection regions ARn (AR1 to AR4) of the above, and in order to set an illumination range suitable for each of the detection regions ADn and ARn. The illumination field diaphragm FAn is provided. As shown in FIG. 9B, the illumination field diaphragm FAn is composed of, for example, three types of illumination field diaphragms FA1, FA2, and FA3. A typical illumination field diaphragm FA1 has a similar shape to both the detection areas ADn and ARn so that the entire detection area ADn (AD1 to AD4) and the detection area ARn (AR1 to AR4) are simultaneously illuminated by the illumination light ILB. Has a rectangular opening.

In FIG. 9A, the illumination light ILb transmitted through the opening of the illumination field diaphragm FA1 is reflected by the cube-type beam splitter BS2, and is split into amplitude by the beam splitter BS1 through the lens system Gb and the objective lens system OBL, and is used as a substrate. The inside of the detection region ADn on P is illuminated with a uniform intensity distribution, and the inside of the detection region ARn on the reference surface RBa of the reference bar member RB is illuminated with a uniform intensity distribution. Since the reference marks RMn (RM1 to RM4) are constantly formed at predetermined positions in the detection region ARn on the reference surface RBa, the reflected light from the detection region ARn (normally reflected light from the reference mark RMn, (Including scattered light and diffracted light) passes through the beam splitter BS1 and then passes through the objective lens system OBL, the lens system Gb, and the beam splitter BS2 to become an imaging light beam Bma and is incident on the imaging unit IMS. The image pickup unit IMS includes a lens system Gc and a two-dimensional image pickup element DIS, and the image pickup surface of the image pickup element DIS includes the surface of the substrate P (or the outer peripheral surface of the rotating drum DR) and the reference surface RBa of the reference bar member RB. It is set to an optically conjugate relationship (imaging relationship) with each of the above.

As shown in FIG. 9C, the image pickup region DIS'on the image pickup surface of the image pickup device DIS is substantially similar to the detection regions ADn and ARn, and has a circular shape defined by the objective lens system OBL and the lens system Gb. It is set to a rectangle included in the image side viewing range Imcc. The reflected light from the detection region ARn forms a magnified image RNn'of the reference mark RNn at a specific position in the imaging region DIS'. In the present embodiment, the reference mark RMn formed on the reference surface RBa of the reference bar member RB is an L-shaped linear pattern arranged at each of the four corners in the detection region ARn (imaging region DIS'). It is composed. In FIG. 9C, the intersection of the line that divides the imaging region DIS'in the vertical direction and the line that divides the imaging region DIS'in the horizontal direction is defined as the center point CCn. The center point CCn may be a specific imaging pixel located at the center of the imaging surface.

On the other hand, when the alignment mark MKn appears in the detection area ADn while the inside of the detection area ADn on the substrate P is illuminated with a uniform intensity distribution by the illumination light ILb transmitted through the opening of the illumination field aperture FA1. After the reflected light including normal reflected light, scattered light, and diffracted light from the alignment mark MKn passes through the beam splitter BS1, it passes through the objective lens system OBL, the lens system Gb, and the beam splitter BS2 to form an imaging light beam Bma and is imaged. It is incident on the part IMS. Since the image pickup surface of the image pickup element DIS of the image pickup unit IMS and the surface of the substrate P are in an imaging relationship, an enlarged image MKn'of the alignment mark MKn is formed in the image pickup region DIS'of the image pickup element DIS. .. Since the substrate P continues to move in one direction at a constant speed, the magnified image MKn'of the alignment mark MKn passes through the imaging region DIS'in the direction of the arrow Xp (X direction). Further, the position of the magnified image RNn'of the reference mark RNn and the position of the magnified image MKn'of the alignment mark MKn in the imaging region DIS'are set so as not to overlap each other. In FIG. 9C, the shape of the alignment mark MKn (enlarged image MKn') is made into a cross-shaped line pattern in accordance with the shape of the alignment mark MKn in FIG. 7, and the shape is a video signal from the image pickup element DIS. Any shape such as a rectangle (square), a triangle (wedge shape), and a circle may be used as long as the shape can be recognized by an image processing device that analyzes the image.

Further, when the illumination field diaphragm FAn of the illumination system ILU is switched to the illumination field diaphragm FA2 shown in FIG. 9B, the reference marks located at the four corners in the detection area ARn set on the reference surface RBa of the reference bar member RB. The illumination light toward RMn is shielded, and only the enlarged image MKn'of the alignment mark MKn on the substrate P appears in the image pickup region DIS'of the image pickup element DIS. Further, when the illumination field diaphragm FAn of the illumination system ILU is switched to the illumination field diaphragm FA3 shown in FIG. 9B, the alignment mark MKn is directed to a portion where the alignment mark MKn can pass in the central portion in the detection region ADn set on the substrate P. Since the illumination light is shielded and the reference mark RMn located at the four corners of the detection region ARn on the reference surface RBa is irradiated with the illumination light, the reference mark RNn is included in the image pickup region DIS'of the image pickup device DIS. Only the magnified image RNn'appears.

FIG. 10A is a diagram showing an example of arrangement of reference marks RM1 to RM4 (RM3 is omitted) formed at four locations in the Y direction on the reference surface RBa of the reference bar member RB. In FIG. 10A, the plane parallel to the reference plane RBa is the X'Y'plane of the Cartesian coordinate system X'Y'Z', and the axis parallel to the normal of the reference plane RBa is the Z'axis. Again, the Y'axis of the Cartesian coordinate system X'Y'Z'is parallel to the Y axis of the Cartesian coordinate system XYZ. Reference marks RM1 to RM4 are formed at predetermined intervals in the Y'direction along a virtual straight line CRy extending in the Y'direction (Y direction) on the reference surface RBa of the reference bar member RB. .. That is, the center points CR1, CR2, CR3, and CR4 of the reference marks RM1 to RM4 are precisely positioned on the virtual straight line CRy. In the present embodiment, assuming that the distance dimension between the center point CR1 of the reference mark RM1 and the center point CR2 of the reference mark RM2 in the Y'direction (Y direction) is LBS12, the center point CR3 and the center point of the other reference mark RM3 Both the distance dimension LBS23 in the Y'direction (Y direction) from CR2 and the distance dimension LBS34 in the Y'direction (Y direction) between the center point CR4 and the center point CR3 of the reference mark RM4 are equal to the distance dimension LBS12. It shall be set. However, the interval dimension LBS12, the interval dimension LBS23, and the interval dimension LBS34 may be set to different values.

FIG. 10B exaggerates an example of the arrangement relationship between the image pickup region DIS'of the image pickup element DIS of the alignment system ALG1 and the reference mark RM1 on the reference bar member RB in the XY'plane. FIG. 10C exaggerates an example of the arrangement relationship between the image pickup region DIS'of the image pickup element DIS of the alignment system ALG2 and the reference mark RM2 on the reference bar member RB in the XY'plane. .. In FIG. 10B, assuming that the center point (reference point) between the X'direction and the Y'direction of the two-dimensional imaging region DIS'is CC1, the reference is due to the relative mounting error between the reference bar member RB and the alignment system ALG1. The center point CR1 in the X'and Y'directions of the mark RM1 and the center point CC1 in the imaging region DIS'are deviated by a predetermined installation error ΔC1. In FIG. 10B, the installation error ΔC1 is + ΔXC1 (μm) in the X'direction and + ΔYC1 (μm) in the Y'direction with reference to the center point CR1 of the reference mark RM1 as a reference (origin).

Similarly, in FIG. 10C, assuming that the center point (reference point) between the X'direction and the Y'direction of the two-dimensional imaging region DIS'is CC2, the relative mounting error between the reference bar member RB and the alignment system ALG2. As a result, the center point CR2 in the X'and Y'directions of the reference mark RM2 and the center point CC2 in the imaging region DIS'are deviated by a predetermined installation error ΔC2. In FIG. 10C, the installation error ΔC2 is −ΔXC2 (μm) in the X'direction and −ΔYC2 (μm) in the Y'direction with reference to the center point CR2 of the reference mark RM2 as a reference (origin). The center points CC1 and CC2 of the imaging region DIS'correspond to a specific imaging pixel located in the center of a large number of imaging pixels distributed in a two-dimensional matrix on the imaging surface. It does not have to be exactly the true center point of the imaging region DIS', for example, the position of a specific imaging pixel deviated from the true center point by several to a dozen in the X'direction or Y'direction. The center points (reference points) CC1 and CC2 may be used.

Similarly, for each of the other alignment systems ALG3 and ALG4, the installation error ΔC3 between each of the center points CC3 and CC4 of the imaging region DIS'and the center points CR3 and CR4 of the reference marks RM3 and RM5, It is assumed that there is ΔC4. Information on these installation errors ΔC1, ΔC2, ΔC3, and ΔC4 can be obtained during equipment assembly, adjustment (calibration) work performed at an appropriate timing during equipment operation, or equipment maintenance inspection (maintenance) work. , It is obtained by image analysis of the video signal from each of the image pickup devices DIS of the alignment systems ALG1 to ALG4, and can be updated as appropriate.

FIG. 11 is a block diagram showing a schematic configuration of a part of a control device provided in the pattern drawing device EX according to the present embodiment. In FIG. 11, the same members and parts as those described in the above drawings are designated by the same reference numerals, but are engraved on the outer peripheral surfaces DRs of the rotating drum DR and the outer peripheral surfaces of the scale disk SDa (SDb). The scale Gm is shown in a flat state for the sake of simplicity. Therefore, in FIG. 11, it is assumed that the outer peripheral surface DRs and the scale Gm of the rotating drum DR go straight together in the horizontal direction (X direction) as shown by the arrow arx. Further, reference patterns FMa, FMb, FMc ... Are formed on the outer peripheral surface DRs of the rotating drum DR at regular intervals along the circumferential direction. Each of the reference patterns FMa, FMb, FMc ... Appears in the detection region ADn (AD1 to AD4) of each of the alignment systems ALGn (ALG1 to ALG4) in the Y direction (central axis AXo of the rotating drum DR). The position and dimensions of the direction) are set and formed on the outer peripheral surfaces DRs of the rotating drum DR. The configuration of the reference patterns FMa, FMb, FMc ... Formed on the rotating drum DR is disclosed in, for example, International Publication No. 2014/034161.

Note that FIG. 11 shows only the odd-numbered drawing units U1, U3, and U5 axially supported by the support frame portion 100, and the encoder heads EHa1 (EHb1) and EHa2 (EHb2) shown in FIGS. 7 and 8 above. , EHa3 (EHb3), the orientation in the circumferential direction of the detection region ADn of the alignment system ALGn and the orientation in the circumferential direction of the drawing lines SL1, SL3, SL5 (exposure position) of the odd-numbered drawing units U1, U3, U5. Only the encoder heads EHa1 (EHb1) and EHa2 (EHb2) to be arranged are shown. Further, as shown in FIG. 8, the reference bar member RB is firmly coupled to the support frame portion 100 that supports the drawing units U1 to U6 via the

support plate portions

102A, 102B, 103A, and 103B.

The measurement signals (for example, two-phase signals having a phase of 90 degrees) from each of the encoder heads EHa1 (EHb1) are input to the counter circuit unit 200A, and the counter circuit unit 200A has 1/32 to 1/32 of the lattice pitch of the scale Gm. Digital counter values corresponding to the movement amount of the scale Gm (that is, the movement amount of the outer peripheral surface DRs of the rotating drum DR or the substrate P) with a resolution of 1/128 are sequentially output as measurement information ESa1 (ESb1). Similarly, measurement signals (for example, two-phase signals having a phase difference of 90 degrees) from each of the encoder heads EHa2 (EHb2) are input to the counter circuit unit 200B, and the counter circuit unit 200B has a grid pitch of the scale Gm (for example). A digital counter value corresponding to the movement amount of the scale Gm (that is, the movement amount of the outer peripheral surface DRs of the rotating drum DR or the movement amount of the substrate P) with a resolution of 1/32, 1/64, or 1/128 of (20.48 μm). It is sequentially output as measurement information ESa2 (ESb2). Further, on each scale Gm of the scale disks SDa and SDb, a zero point signal indicating the starting point (zero point position) is transmitted to each of the encoder heads EHa1 (EHb1), EHa2 (EHb2), and EHa3 (EHb3). The origin pattern ZZo for generating from is formed.

The counter circuit unit 200A shown in FIG. 11 resets the digital count value to zero at the moment when the encoder head EHa1 (EHb1) detects the origin pattern ZZo, and the counter circuit unit 200B has the encoder head EHa2 (EHb2) resetting the digital count value to zero. The digital count value is reset to zero the moment it is detected. The origin pattern ZZo on the scale disk SDa side and the origin pattern ZZo on the scale disk SDb side do not necessarily have to be exactly the same direction with respect to the rotation direction (circumferential direction) of the rotating drum DR, and are substantially the same direction, for example. The angle difference may be set within the range of ± several degrees. As an example, when an optical rotary encoder system manufactured by Renishaw is used, a scale disk SDa in which a scale scale Gm is engraved on the outer peripheral surface at a pitch of 20 μm, a stainless steel ring as SDb and an encoder head are used to rotate the scale scale Gm. The amount of movement in the direction can be measured with a resolution of 0.1 μm or less. That is, the change of the least significant bit (LSB) of the digital counter in the

counter circuit units

200A and 200B can correspond to the movement amount of the substrate P by about 0.1 μm.

Assuming that the radius of the outer peripheral surface on which the scale scale Gm of the scale disks SDa and SDb is engraved is φgm and the radius of the outer peripheral surface of the rotating drum DR is φdr, the scale scale Gm is directly applied to the outer peripheral surface of the rotating drum DR. Unless it is engraved, it is difficult to make the radius φgm and the radius φdr exactly the same. Therefore, in the

counter circuit units

200A and 200B, the unit movement amount corresponding to the LSB of the counter value that is first digitally counted based on the signals from the encoder heads EHa1, EHb1, EHa2, EHb2, ... When ΔLx is used as the measurement resolution), a function is provided for converting the unit movement amount ΔLxa in the circumferential direction of the outer peripheral surface of the rotating drum DR by the calculation of ΔLxa = ΔLx · (φdr / φgm). Therefore, the measurement information ESa1, ESb1, ESa2, ESb2, ... Output from each of the

counter circuit units

200A and 200B shown in FIG. 11 are obtained by such conversion in the circumferential direction of the outer peripheral surface of the rotating drum DR. It can be a value that directly represents the amount of movement (or the position of movement).

Further, since it is necessary to accurately measure the movement amount and the movement position in the circumferential direction of the surface of the substrate P curved in a cylindrical surface, the

counter circuit units

200A and 200B are inside the

counter circuit units

200A and 200B based on the thickness Tp of the substrate P. A function of converting the unit movement amount ΔLxb in the circumferential direction of the surface of the substrate P by the calculation of ΔLxb = ΔLx · [(φdr + Tp) / φgm] is also provided, and the

counter circuit units

200A and 200B are respectively measured information ESa1 and ESb1. , ESa2, ESb2, ..., In addition to the unit movement amount ΔLxa, the movement amount and the movement position measured based on the unit movement amount ΔLxb are simultaneously output. As an example, the radius φdr of the outer peripheral surface of the rotating drum DR is 134.00 mm, the thickness Tp of the substrate P is 100 μm, and the amount of movement of the outer peripheral surface of the rotating drum DR (of the outer peripheral surface) when the rotating drum DR is rotated exactly once. Let's compare the total circumference (long distance) with the amount of movement of the surface of the substrate P.

The amount of movement of the outer peripheral surface of the rotating drum DR (total circumference long distance) is 841.946 mm [= 2π · φdr], and the amount of movement of the surface of the substrate P is 848.229 mm [= 2π · (φdr + Tp)], which is the difference. (Error length) ΔLf is about 628.3 μm. The error length ΔLf has a large error of + 0.746% [= ΔLf / (2π · φdr)] with respect to the movement amount measured based on the unit movement amount ΔLxa of the outer peripheral surface of the rotating drum DR. Become. Therefore, in the state before the substrate P is passed through the rotary drum DR, at the stage of calibration or the like for detecting the reference patterns FMa, FMb, FMc ... On the outer peripheral surface of the rotary drum DR, the rotary drum The movement amount and movement position measured based on the unit movement amount ΔLxa corresponding to the outer peripheral surface of the DR are used, and the pattern is exposed on the substrate P in the state after the substrate P is passed through the rotating drum DR. Or, at the stage of detecting the alignment marks MKn (MK1 to MK4), the movement amount and the movement position measured based on the unit movement amount ΔLxb corresponding to the surface of the substrate P are used.

Each photoelectric signal (analog voltage) from the photoelectric sensor DTR provided in each of the odd-numbered drawing units U1, U3, and U5 is input to the ADC unit 202A including three analog / digital conversion circuits. The ADC unit 202A measures the intensity of the photoelectric signal from the photoelectric sensor DTR of the drawing unit U1 (U3, U5) in the Y direction along the drawing line SL1 (SL3, SL5), and the unit scanning movement amount ΔYsp of the spot light. Digital sampling is performed based on the clock signal LTC that generates a clock pulse every (μm). The unit scanning movement amount ΔYsp is a fiber amplifier laser light source in which the light source device LS shown in FIG. 1 pulsates at a frequency FPL (Hz), and spot light for each continuous pulse oscillation spots in the main scanning direction (Y direction). When the scanning speed Vsp (μm / sec) of the spot light is set so as to overlap at 1/2 or more of the diameter of the light, the relationship of ΔYsp ≧ Vsp / FPL is set. In this relationship, when ΔYsp = Vsp / FPL, the unit scanning movement amount ΔYsp is about ½ of the diameter of the spot light, and the frequency of the clock signal LTC is the same as the frequency FPL of the light source device LS. Further, when the emission clock signal of the frequency FPL for pulse oscillation of the light source device LS is reduced to 1/2 frequency by a frequency divider circuit or the like and generated as a clock signal LTC, the relationship is ΔYsp = 2 · Vsp / FPL. The unit scanning movement amount ΔYsp is the same as the diameter of the spot light.

The ADC unit 202A sequentially transfers a large number of digital data digitally sampled for each unit scanning movement amount ΔYsp during one scanning of the spot light along the drawing line SL1 (SL3, SL5) to the image processing unit 204A. The image processing unit 204A sequentially and individually stores the digital waveform data of the photoelectric signals from the photoelectric sensor DTRs of the odd-numbered drawing units U1, U3, and U5 digitally sampled by the ADC unit 202A in the internal image memory. .. At that time, the image memory sequentially updates the storage address of the digital data obtained during one scan of the spot light based on the change of the measurement information ESa2 (or ESb2) from the counter circuit unit 200B.

As a result, in the image memory in the image processing unit 204A, the reference patterns FMa, FMb, formed on the outer peripheral surface DRs of the rotating drum DR within the scanning range of the spot light along each of the drawing lines SL1, SL3, and SL5. Two-dimensional light-dark image data (luminance information) is generated by the reflected light from FMc ... Etc. and the reflected light from fine pattern shapes having different reflectances formed on the substrate P.

Similarly, for each of the even-numbered drawing units U2, U4, and U6, an ADC unit 202B (not shown) that converts the photoelectric signal from the photoelectric sensor DTR into digital waveform data and an even-numbered drawing in the image memory. An image processing unit 204B (not shown) for individually storing the digital waveform data of the photoelectric signal from each of the photoelectric sensor DTRs of the units U2, U4, and U6 is provided. The digital sampling of the photoelectric signal from each photoelectric sensor DTR by the ADC unit 202B is performed based on the clock signal LTC in FIG. 11, but the digital waveform data to be temporarily stored in the image memory in the image processing unit 204B The storage address is not the measurement information ESa2 (or ESb2) from the counter circuit unit 200B in FIG. 11, but the measurement signal from the encoder head EHa3 (or EHb3) in FIG. 8A (for example, two phases having a phase difference of 90 degrees). It is sequentially updated based on the change of the measurement information (referred to as ESa3 or ESb3) from the counter circuit unit 200C (not shown) that counts the signal). The counter circuit unit 200C (not shown) also resets the digital count value to zero at the moment when the encoder head EHa3 (EHb3) detects the origin pattern ZZo.

In the above image processing unit 204A (and 204B), the position (pixel position) in the image with respect to the main scanning direction (Y direction) of the two-dimensional image data stored in the image memory is a clock signal from the start of digital sampling. The position (pixel position) in the image with respect to the sub-scanning direction (X direction), which is determined by the count value of the clock pulse of the LTC, is the measurement information ESa2, ESb2 (and ESa3, ESb3) from the counter circuit unit 200B (and 200C). Determined by. As a result, the image processing unit 204A (and 204B) has the two-dimensional positions of the reference patterns FMa, FMb, FMc ... Of the rotating drum DR based on the two-dimensional image data stored in the image memory. That is, the relative positions of the reference patterns FMa, FMb, FMc ... With respect to the drawing lines SLn (SL1 to SL6) by the spot light are measured by the image analysis process.

Alignment system ALGn (ALG1 to ALG4) that detects reference patterns FMa, FMb, FMc ... of the rotating drum DR, reference marks RM1 to RM4 of the reference bar member RB, or alignment marks MKn (MK1 to MK4) on the substrate P. The video signal Vsg from each image sensor DIS is sent to the image analysis unit 206. The image analysis unit 206 is used to capture image data of the image pickup region DIS'that is imaged by each image sensor DIS and sent sequentially (for example, data for one screen that is refreshed every 1/30 second or 1/60 second). ) Is temporarily stored in the image memory at the moment when the measurement information ESa1 (or ESb1) from the counter circuit unit 200A reaches the measured value corresponding to the specified trigger position. The trigger position is the rotating drum DR such that when the imaging target is the reference pattern FMa, FMb, FMc ..., The reference pattern FMa, FMb, FMc ... Is located substantially in the center of the imaging area DIS'. It is set by storing the measurement information ESa1 (or ESb1) output by the counter circuit unit 200A at the rotation position in advance.

Further, as shown in FIG. 7, when the alignment systems ALG1 and ALG4 target the alignment marks MK1 and MK4 on the substrate P for imaging, the intervals (for example, 5) related to the sub-scanning direction of the alignment marks MK1 and MK4 are designed. Since (~ 20 mm) is fixed, the trigger position is set for each rotation position of the rotating drum DR in which the measurement information ESa1 (or ESb1) from the counter circuit unit 200A is increased by the count value corresponding to the interval value. NS.

When the image analysis unit 206 measures the position of the alignment mark MKn by the alignment system ALGn, the image analysis unit 206 is centered in the imaging region DIS'as shown in FIG. 9C above based on the image data stored in the image memory. Alignment mark MKn with respect to point CCn Alignment with respect to the two-dimensional displacement of the center point of MKn'or the magnified image RNn of the reference bar member RB that appears simultaneously in the imaging region DIS' The amount of two-dimensional misalignment of the center point of the magnified image MKn'of the mark MKn is measured by image analysis processing. Further, when the image analysis unit 206 measures the positions of the reference patterns FMa, FMb, FMc ... Of the rotating drum DR by the alignment system ALGn, the center in the imaging region DIS'as in FIG. 9C above. The two-dimensional displacement of the center point of the enlarged image of the reference patterns FMa, FMb, FMc ... With respect to the point CCn, or the reference mark RMn of the reference bar member RB that appears simultaneously in the imaging region DIS'. The amount of two-dimensional misalignment of the center point of the magnified image of the reference patterns FMa, FMb, FMc ... With respect to the magnified image RMn'is measured by image analysis processing.

Further, as described in FIGS. 10B and 10C above, the image analysis unit 206 sets each of the center points CCn (CC1 to CC4) in each imaging region DIS'of the alignment system ALGn and the reference mark of the reference bar member RB. Information on the installation error ΔCn (ΔC1 to ΔC4) with each of the center points CRn (CR1 to CR4) of the RMn is measured by image analysis processing of the image data stored in the image memory.

The measurement control unit (calculation processing unit) 210 of FIG. 11 is the relative position of the reference patterns FMa, FMb, FMc ... With respect to the drawing lines SLn (SL1 to SL6) measured by the image processing unit 204A (204B). Spots along each of the drawing lines SLn based on the position of the alignment mark MKn detected by the alignment system ALGn based on the relationship information and the information of various misalignment amounts measured by the image analysis unit 206. The amount of adjustment of drawing timing by light (setting of delay time), the amount of adjustment of each inclination of drawing line SLn by each minute rotation of drawing unit Un, and the amount of adjustment of drawing line SLn itself by a small amount in the sub-scanning direction. The amount of adjustment of the inclination of the parallel flat plate HVP (see FIG. 3) is calculated at high speed. In calculating the adjustment amount, the measurement control unit (calculation processing unit) 210 uses the position information (actual measurement information) of the alignment marks MKn (MK1 to MK4) of the substrate P detected by the alignment system ALGn (ALG1 to ALG4). ) Uses the correction position information corrected according to the installation error information ΔC1 to ΔC4 of the alignment system ALGn (ALG1 to ALG4) described with reference to FIG.

[Operation sequence]
FIG. 12 is a flowchart illustrating an example of a series of operations of a calibration sequence, an alignment sequence, and an exposure sequence according to the first embodiment. In FIG. 12,

steps

300, 302, and 304 are various calibration sequences executed at the time when the device is started up or at an appropriate time during the operation of the device, and step 306 is a transfer mechanism (various rollers) in the device. This is a paper-passing operation in which a substrate P having a photosensitive functional layer formed on a rotating drum DR is hung with a predetermined tension. Steps 308 to 314 of FIG. 12 are first exposure (1st exposure) sequences for exposing the pattern for the first layer to the photosensitive functional layer of the passed substrate P, and steps 316, 318, 320, and 322, Reference numeral 324 is a second exposure (2nd exposure) sequence in which the patterns of the second and subsequent layers are superimposed and exposed on the photosensitive functional layer of the substrate P on the pattern of the first layer, and an alignment sequence for the overlay.

Further, in the present embodiment, as shown in FIG. 13, it is assumed that the exposed region DPAs for a plurality of display panels are arranged over the entire length of the substrate P supplied from one roll in the long direction. FIG. 13 shows a state in which the substrate P is stretched on a flat surface. The substrate P is passed through the rotating drum of the pattern drawing apparatus EX in the direction of the arrow arc from the tip Pa, and the rotating drum DR rotates to the terminal Pb. Be transported. The first region F0 from the tip Pa of the substrate P is a margin portion until the substrate P is passed through the pattern drawing apparatus EX and can be conveyed at a predetermined tension and a predetermined speed. In the next region F1 on the substrate P, four alignment marks MK1 to MK4 are arranged in the width direction of the substrate P so that they can be detected in the detection regions AD1 to AD4 of the four alignment systems ALG1 to ALG4 shown in FIG. Is placed in. A plurality of rows of the four alignment marks MK1 to MK4 arranged in the width direction are formed at regular intervals (for example, 5 mm to 10 mm) in the long direction. In the next regions F2 to Fn-1 on the substrate P, a large number of exposed region DPA and alignment marks MK1 to MK4 attached to the exposed region DPA are formed in the long direction with a fixed length margin portion interposed therebetween. The last region Fn on the substrate P is a margin portion where the exposed region DPA is not formed because normal tension cannot be maintained even when the substrate P is hung around the rotating drum DR.

The alignment marks MK1 to MK4 shown in FIG. 13 are simultaneously exposed at the time of the 1st exposure process of exposing the pattern of the first layer to the exposed area DPA, and the etching process and the plating process performed after the exposure process. It is formed together with the pattern for the first layer by another film forming process. Therefore, a metal thin film (copper, aluminum, nickel) or an opaque insulating film as a pattern for the first layer is formed on the entire surface of the substrate P to be subjected to the 1st exposure treatment, and further, the metal thin film (copper, aluminum, nickel) or an opaque insulating film is formed on the entire surface. It is preferable to form a photosensitive functional layer on the surface of a metal thin film or an insulating film.

[Step 300]
Returning to the description of the flowchart of FIG. 12, the encoder measurement system is calibrated in step 300 in the state after the device is started and before the paper is passed. In FIG. 12, [DR] added next to step 300 means that the rotary drum DR is rotationally driven at a predetermined rotation speed, and [EH] is the encoder heads EHa1 to EHa3, EHb1 to EHb1 of the encoder measurement system. It means to use each of EHb3. As shown in FIGS. 7 and 11 above, the outer peripheral surfaces (formation surface of the scale Gm) of the scale disks SDa and SDb at both ends of the rotary drum DR have a perfect circle error (at the rotation angle position) determined by the machining accuracy. It includes an error in which the radius from the central axis AXo changes slightly from the design value) and a slight eccentricity error at the time of mounting with respect to the central axis AXo. Further, a slight unevenness (pitch error) may occur in the pitch in the circumferential direction of the scale Gm engraved on the outer peripheral surfaces of the scale disks SDa and SDb.

In step 300, the perfect circle error, the eccentric error, and the pitch error are accurately grasped, and an error correction map at the time of encoder measurement is created. Methods for obtaining these errors and preparation of a correction map are disclosed in, for example, International Publication No. 2016/013417 and JP-A-2017-090243. The error correction map starts from the origin pattern ZZo shown in FIG. 11 and divides one round (360 degrees) of the scale disks SDa and SDb at a fixed angle (for example, 5 degrees) as an error correction amount for each rotation position. It will be remembered. The angular position information (movement amount) of the scale Gm actually measured by each of the encoder heads EHa1 to EHa3 (EHb1 to EHb3) is the counter circuit unit (counter circuit unit 200A in FIG. 11) prepared for each head. 200B etc.). Therefore, each counter circuit unit (200A, 200B, etc.) outputs the value obtained by correcting the measured value by the created error correction map (error correction amount) as measurement information (ESa1, ESb1, ESa2, ESb2, etc. in FIG. 11). do. When the calibration of the encoder measurement system is completed, the calibration of the alignment system is executed in the next step 302.

[Step 302]
In FIG. 12, [DR] and [EH] added next to step 302 are the rotational drive of the rotating drum DR and the measurement by the encoder measurement system (encoder heads EHa1 to EHa3, EHb1 to EHb3) as in step 300. [RB] means to use the reference bar member RB shown in FIGS. 4 to 6 and 8 to 11, and [ALGn] means to use the reference bar member RB shown in FIGS. 4 to 6 and 8 to 11. 6, It means that the four alignment systems ALG1 to ALG4 shown in FIGS. 9 and 11 are used.

In step 302, each of the four alignment systems ALG1 to ALG4 measures each position of the corresponding reference marks RM1 to RM4 on the reference bar member RB shown in FIG. 10A by the image analysis unit 206 in FIG. As described with reference to FIGS. 10B and 10C, there is an installation error ΔC1 between the center point CC1 of the (detection area AD1) imaging area DIS'(detection area AD1) of the alignment system ALG1 and the center point CR1 of the reference mark RM1, and the alignment system ALG2. Installation error ΔC2 between the center point CC2 of the imaging region DIS'(detection region AD2) and the center point CR2 of the reference mark RM2, and the center point CC3 and the reference mark RM3 of the imaging region DIS'(detection region AD3) of the alignment system ALG3. The installation error ΔC3 with the center point CR3 and the installation error ΔC4 between the center point CC4 of the imaging region DIS'(detection region AD4) of the alignment system ALG4 and the center point CR4 of the reference mark RM4 are measured. At that time, when the rotating drum DR is rotated, the images of the reference patterns FMa, FMb, FMc ... On the outer peripheral surface DRs also appear in the respective imaging regions DIS'of the alignment systems ALG1 to ALG4. The illumination field diaphragm FAan provided in each of the illumination systems ILU of the alignment systems ALG1 to ALG4 shown in FIG. 9A may be switched to the illumination field diaphragm FA3 shown in FIG. 9B.

From the above sequence, as schematically shown in FIG. 14, information on the installation errors ΔC1 to ΔC4 of each of the four alignment systems ALG1 to ALG4 can be obtained. FIG. 14 is an exaggerated representation of the installation errors ΔC1 to ΔC4 based on the center points CR1 to CR4 of the reference marks RM1 to RM4 on the reference bar member RB shown in FIG. 10A. The mutual placement accuracy (interval error) of each center point CRn (n = 1 to 4) of each reference mark RNn (n = 1 to 4) on the reference bar member RB is a processing device (patterning) for forming the reference mark RMn. Although it depends on the positioning accuracy of the device), if the processing device is equipped with a positioning stage or the like equipped with a laser interferometer, the placement accuracy (placement error) with respect to the design position can be set to ± 0.2 μm or less. Therefore, the installation error information ΔCn (n = 1 to 4) measured by the image analysis unit 206 corresponds to the installation error from each design position of the alignment system ALGn, and is about ± several μm, if it is large. It will be about a dozen μm.

In step 302, by detecting the images of the reference patterns FMa, FMb, FMc ... Formed on the outer peripheral surfaces DRs of the rotating drum DR by the alignment system ALGn, the Y'direction on the reference bar member RB ( It is possible to confirm the parallelism between the straight line CRy extending in the Y direction) and the rotation center axis AXo of the rotation drum DR. FIG. 15 shows the arrangement relationship between the reference patterns FMa, FMb, FMc ... Formed on the outer peripheral surfaces DRs of the rotating drum DR developed in a plane and the detection regions AD1 to AD4 of the alignment systems ALG1 to ALG4. It is a figure which shows an example. In FIG. 15, reference patterns FMa, FMb, FMc ... Are provided at eight positions at 45 ° degrees along the circumferential direction on the outer peripheral surface DRs. Each of the reference patterns FMa to FMh extends in the circumferential direction orthogonal to the linear pattern Fyo and the linear pattern Fyo engraved linearly in the Y direction parallel to the central axis (center line) AXo of the rotation of the rotating drum DR. Therefore, it is composed of linear patterns Fx1, Fx2, Fx3, and Fx4 that are engraved corresponding to the positions of the detection regions AD1 to AD4 in the Y direction. The spacing between the four linear patterns Fx1 to Fx4 in the Y direction is set to be the same as the spacing dimensions LBS12, LBS23, and LBS34 of the reference marks RM1 to RM4 shown in FIG. 10A.

With the rotation of the rotating drum DR, for example, in the detection area AD1 of the alignment system ALG1 (in the imaging area DIS'), the intersection portion of each of the linear patterns Fyo and the linear pattern Fx1 of the reference patterns FMa to FMh is formed. , Appears 8 times during one rotation of the rotating drum DR. The timing at which each of the reference patterns FMa to FMh appears in the detection region ADn of the alignment system ALGn is the circumferential position of the origin pattern ZZo of the scale Gm of the scale disk SDa and SDb and the circumferential position of each of the reference patterns FMa to FMh. Since the relationship with is fixed in advance (known), it can be specified from the rotation angle position of the rotation drum DR measured by the encoder measurement system. Therefore, the alignment system is based on the measurement information ESa1 (ESb1) from the counter circuit unit 200A in FIG. 11 for inputting the measurement signal from the encoder head EHa1 (EHb1) arranged corresponding to the circumferential position of the alignment system ALGn. At the timing when each linear pattern Fyo of the reference patterns FMa to FMh appears in each detection region ADn of ALGn (the timing when the measurement information ESa1 becomes a specific measurement value), the reference mark RMn of the reference bar member RB ( The image of n = 1 to 4) and the image of the intersection of the linear pattern Fyo and the streak pattern Fxn (n = 1 to 4) are simultaneously sampled by each of the four image pickup elements DIS, and are simultaneously sampled by the image analysis unit 206. , The misalignment amounts ΔCrf1, ΔCrf2, ΔCrf3, and ΔCrf4 between the center point Cfm of the intersection and the center point CRn of the reference mark RNn (n = 1 to 4) are measured. At that time, the illumination field diaphragm FAan provided in each illumination system ILU of the alignment system ALGn shown in FIG. 9A is switched to the illumination field diaphragm FA1 shown in FIG. 9B.

The measurement control unit (calculation processing unit) 210 in FIG. 11 has an outer peripheral surface measured at each position of the four detection regions ADn in the position deviation amount ΔCrfn (n = 1 to 4) from the image analysis unit 206. Based on the displacement amount component in the circumferential direction of the DRs, the relative inclination error Δθrb of each of the linear patterns Fyo of the reference patterns FMa to FMh and the linear CRy set on the reference bar member RB is calculated. The calculated tilt error Δθrb is indirectly regarded as a tilt error (parallelism error) in the circumferential direction between the center line AXo of rotation of the rotating drum DR and the straight line CRy on the reference bar member RB. Since the tilt error Δθrb can be measured for each position of the reference patterns FMa to FMh at eight locations on the outer peripheral surface DRs (every 45 ° rotation of the rotating drum DR), the tilt error Δθrb for eight times is averaged. Is also good.

The measured tilt error Δθrb is set to be within the permissible range during normal operation of the equipment, but when the equipment is restarted after a long suspension of operation or when it is restarted after receiving a large vibration due to an earthquake or the like. , The tilt error Δθrb may exceed the permissible range. Therefore, when the inclination error Δθrb is measured as described above and exceeds the permissible range, the inclination of the reference bar member RB can be adjusted by the fine adjustment mechanism 106 shown in FIG. However, when the inclination of the reference bar member RB is adjusted by the fine adjustment mechanism 106, the alignment system calibration sequence of step 302 is executed again, and the mounting error of the alignment system ALGn from each design position causes the alignment error. Certain installation error information ΔC1 to ΔC4 are measured again.

Further, in step 302, since each of the four alignment systems ALGn (n = 1 to 4) can sequentially detect the reference patterns FMa to FMh as the rotating drum DR rotates, each of them as shown in FIG. The relative positional relationship of each of the sub-scanning directions (X'direction in FIG. 10) of the center points CC1 to CC4 of the imaging region DIS'is determined with reference to the reference patterns FMa to FMh. In this case, for example, at the timing when the linear pattern Fyo of the reference pattern FMa appears in the imaging region DIS'of each of the four alignment systems ALGn, the linear pattern Fyo and the linear pattern Fx1 to the linear pattern Fyo and the linear pattern Fx1 to each of the image pickup elements DIS of the alignment system ALGn. The image of the intersection with Fx4 is sampled at the same time, and the value of the measurement information ESa1 (ESb1) output from the counter circuit unit 200A at that time is stored.

By analyzing the sampled image data of the alignment systems ALG1 to ALG4 by the image analysis unit 206 and the measurement control unit 210 in FIG. 11, sub-scanning of the center points CC1 to CC4 with respect to the linear pattern Fyo of the reference pattern FMa. The amount of misalignment with respect to the direction (X'direction in FIG. 10) can be obtained. As a result, the coordinate positions of the center points CC1 to CC4 with respect to the circumferential direction (sub-scanning direction) in the coordinate system of the outer peripheral surface of the rotating drum DR defined by the measurement information ESa1 (ESb1) from the counter circuit unit 200A. , The relative positional relationship (arrangement error) of each of the center points CC1 to CC4 is determined.

[Step 304]
Next, in step 304 of FIG. 12, calibration of the drawing unit Un (n = 1 to 6) is executed. Here, mainly, a drawing position adjustment sequence for measuring and correcting an error in the relative positional relationship of each of the drawing lines SLn (n = 1 to 6) by each of the drawing units Un, and an alignment system ALGn (n = Sub-scanning direction (of the outer peripheral surface DRs of the rotating drum DR) between the center point CCn (n = 1 to 4) of each detection area ADn (imaging area DIS') of each of 1 to 4) and each of the drawing lines SLn. A baseline management sequence for acquiring information about the distance relationship (circumferential direction) or the distance relationship related to the main scanning direction (Y direction) is executed. Note that [DR] and [EH] added next to step 304 in FIG. 12 are the rotational drive of the rotating drum DR and the encoder measurement system (encoder heads EHa1 to EHa3, EHb1 to EHb3), as in steps 300 and 302. ) Is used, and [DTR] means that the photoelectric sensor DTR provided in each of the drawing units Un shown in FIGS. 3 and 11 above is used.

In the drawing position adjustment sequence, while rotating the rotating drum DR at a constant speed, each polygon mirror (rotating multifaceted mirror) PM of the drawing unit Un is rotated at the rotating drum DR (peripheral speed of the outer peripheral surfaces DRs of the rotating drum DR). ) Is rotated at a predetermined rotation speed. When the rotation speed of the rotating drum DR and the rotation speed of the polygon mirror PM are stable, the intensity-modulated drawing beam LBn (n = 1 to 6) based on the dummy drawing data is rotated from each of the drawing units Un. It is projected onto the outer peripheral surface DRs of the drum DR. The dummy drawing data always turns each of the beam LBn (n = 1 to 6) on over the length of each drawing line SLn (n = 1 to 6) which is the scanning locus of the spot light SP in the Y direction. It modulates the state.

As the rotating drum DR rotates, each spot light of the beam LBn projected from each of the drawing units Un sequentially scans each of the reference patterns FMa, FMb, FMc ... Formed on the outer peripheral surface DRs. .. When each spot light scans, for example, a reference pattern FMa (FMb, FMc ... May be used), as described in FIG. 11 above, the photoelectric sensor DTR provided in each of the drawing units Un determines the amount of reflected light. A photoelectric signal corresponding to the change is output, and the ADC unit 202A (202B) and the image processing unit 204A (204B) generate two-dimensional light / dark image data (luminance information) including the reference pattern FMa. Here, the reference pattern FMa (FMb, FMc. The positional relationship of each of the drawing lines SLn (n = 1 to 6) with reference to (may be) is measured. The reflected light monitor system is composed of the photoelectric sensor DTR, the ADC unit 202A (202B), and the image processing unit 204A (204B), and in addition to the bright and dark image data of the reference pattern FMa (FMb, FMc ...). It is also possible to acquire light and dark image data corresponding to the patterns and alignment marks formed on the substrate P.

FIG. 16 shows an example of the arrangement relationship of the imaging region DIS'(detection region ADn) of each of the six drawing lines SLn and the four alignment system ALGn and the reference pattern FMa based on the configuration of FIG. , It is the figure which developed and represented a part of the outer peripheral surface DRs of a rotary drum DR in a plane. Here, the main scanning direction of the spot light forming each of the odd-numbered drawing lines SL1, SL3, and SL5 is set to the −Y direction, and the main scanning direction of the spot light forming each of the even-numbered drawing lines SL2, SL4, and SL6 is set. The direction shall be set in the + Y direction. Further, as shown in FIGS. 2, 3 and 8A above, the midpoints CE1 to CE6 in the main scanning directions (Y direction) of the drawing lines SL1 to SL6 are minute in each of the drawing units U1 to U6. It is arranged so that the extension of the line segments LE1 to LE6, which becomes the rotation center line when rotating, passes through.

The circumferential position CXA in FIG. 16 is the center point CCn (n = 1 to 4) of each of the alignment systems ALGn (n = 1 to 4) measured by the encoder system (counter circuit unit 200A in FIG. 11). ) Is determined by averaging the position information in the sub-scanning direction (circumferential direction), or of the alignment systems ALGn (n = 1 to 4), two alignment systems ALG1 and ALG4 installed on both sides in the Y direction. Represents a position determined by averaging the position information in the sub-scanning direction (circumferential direction) of each of the center points CC1 and CC4.

Further, the reference pattern FMa (same for FMb, FMc ...) On the outer peripheral surface DRs of the rotating drum DR shown in FIG. 15 is circumferential so as to intersect the linear pattern Fyo in addition to the linear patterns Fx1 to Fx4. A large number of linear patterns extending in the direction and arranged at each of the plurality of positions in the Y direction are formed as shown in FIG.

As shown in FIG. 16, the streak pattern Fx1 detected in the imaging region DIS'(center point CC1) as the detection region AD1 of the alignment system ALG1 is arranged near the end of scanning within the scanning range of the drawing line SL1. The drawing unit U1 has a linear pattern Fxe1 that can be detected by the reflected light monitor system (photoelectric sensor DTR). The streak pattern Fx2 detected in the imaging region DIS'(center point CC2) as the detection region AD2 of the alignment system ALG2 is arranged near the end of scanning within the scanning range of the drawing line SL2, and the reflected light of the drawing unit U2. The streak pattern Fxe2, which can be detected by the monitor system (photoelectric sensor DTR), and the reflected light monitor system (photoelectric sensor DTR) of the drawing unit U3, which is arranged near the end of scanning within the scanning range of the drawing line SL3, can also be detected. It has a streak pattern Fxe3.

Similarly, the streak pattern Fx3 detected in the imaging region DIS'(center point CC3) as the detection region AD3 of the alignment system ALG3 is arranged near the end of scanning within the scanning range of the drawing line SL4, and the drawing unit U4 The streak pattern Fxe4, which can be detected by the reflected light monitor system (photoelectric sensor DTR) of the above, and the reflected light monitor system (photoelectric sensor DTR) of the drawing unit U5, which is arranged near the end of scanning within the scanning range of the drawing line SL5. It has a detectable streak pattern Fxe5. Further, the streak pattern Fx4 detected in the imaging region DIS'(center point CC4) as the detection region AD4 of the alignment system ALG4 is arranged near the end of scanning within the scanning range of the drawing line SL6, and is arranged in the drawing unit U6. It has a streak pattern Fxe6 that can be detected even by a reflected light monitor system (photoelectric sensor DTR).

Further, the reference pattern FMa is a linear pattern Fxs1 arranged at a position corresponding to the vicinity of the scanning start in the scanning range of the drawing line SL1, and a linear pattern arranged at a position corresponding to the vicinity of the midpoint CE1 of the drawing line SL1. Fxc1, the linear pattern Fxs2 arranged at the position corresponding to the vicinity of the scanning start in the scanning range of the drawing line SL2, the linear pattern Fxx2 arranged at the position corresponding to the vicinity of the midpoint CE2 of the drawing line SL2, and the drawing line SL3. Within the scanning range of the linear pattern Fxs3 arranged at a position corresponding to the vicinity of the scanning start in the scanning range, the linear pattern Fxc3 arranged at a position corresponding to the vicinity of the midpoint CE3 of the drawing line SL3, and the drawing line SL4. Corresponds to the vicinity of the start of scanning within the scanning range of the linear pattern Fxs4 arranged at the position corresponding to the vicinity of the scanning start, the linear pattern Fxc4 arranged at the position corresponding to the vicinity of the midpoint CE4 of the drawing line SL4, and the drawing line SL5. The linear pattern Fxs5 arranged at the designated position, the linear pattern Fxc5 arranged at the position corresponding to the vicinity of the midpoint CE5 of the drawing line SL5, and the position corresponding to the vicinity of the scanning start within the scanning range of the drawing line SL6. It has a linear pattern Fxs6 to be arranged and a linear pattern Fxc6 arranged at a position corresponding to the vicinity of the midpoint CE6 of the drawing line SL6.

Further, the distance between the streak pattern Fxc1 and the streak pattern Fxc2 in the Y direction YJ12, the distance between the streak pattern Fxc2 and the streak pattern Fxc3 in the Y direction YJ23, and the distance between the streak pattern Fxc3 and the streak pattern Fxc4 in the Y direction. The interval YJ34, the interval YJ45 between the linear pattern Fxc4 and the linear pattern Fxc5 in the Y direction, and the interval YJ56 between the linear pattern Fxc5 and the linear pattern Fxc6 in the Y direction are the midpoints of the drawing lines SL1 to SL6. It is set to be a design distance (for example, 52.00 mm) in the Y direction between CE1 to CE6. In the present embodiment, since the scanning lengths of the drawing lines SL1 to SL6 are set to the same value (for example, 52.00 mm), the distances between the midpoints CE1 to CE6 adjacent to each other in the Y direction are also equal. However, the interval is set to be slightly due to the mounting error of the drawing units U1 to U6 at the time of assembling the device, the mounting error of the optical member in each drawing unit Un, the change in the environmental temperature, etc. With the error of.

Therefore, in the drawing position adjustment sequence in step 304, the reference pattern FMa (other reference patterns FMb, FMc ... May be used) on the rotating drum DR is determined to cross each of the drawing lines SL1 to SL6 in the circumferential direction. While rotating the rotating drum DR at a speed, drawing is performed based on the measurement information ESa2 (ESb2) regarding the moving position of the outer peripheral surface DRs of the rotating drum DR measured by the counter circuit unit 200B corresponding to the encoder head EHa2 (EHb2). The reflected light monitor system (photoelectric sensor DTR) of each of the units U1 to U6 is used to acquire the light / dark image data of the corresponding portion of the reference pattern FMa. Specifically, when the measurement information ESa2 (ESb2) from the counter circuit unit 200B shown in FIG. 11 reaches the circumferential position CX1 in FIG. 16, the odd-numbered drawing units U1, U3, and U5 are respectively. The sampling of the light / dark image data by the reflected light monitor system is started, and the sampling is finished when the measurement information ESa2 (ESb2) reaches the circumferential position CX2. The sampling of light and dark image data by the reflected light monitor system is disclosed in, for example, International Publication No. 2015/152217 and International Publication No. 2018/066285.

In FIG. 16, the circumferential positions CX1 and CX2 represent the moving positions of the reference pattern FMa measured by the counter circuit unit 200B with a resolution of, for example, 0.1 μm, and the drawing lines SL1 and SL3 having odd-numbered reference patterns FMa. It is set in a range that can sufficiently cross SL5. Further, the peripheral positions CX1 and CX2 are the movement amounts (below the unit movement amount ΔLxa) after the digital counter in the counter circuit unit 200B is zero-reset by the origin pattern ZZo in the encoder system using the encoder head EHa2 or EHb2. It is also the circumference measured by). Since the relationship between the origin pattern ZZo and the reference pattern FMa (or other reference patterns FMb, FMc ...) in the circumferential direction, that is, the distance between the origin pattern ZZo and the reference pattern FMa (or other reference patterns FMb, FMc ...) in the circumferential direction on the outer peripheral surface DRs is generally known. , CX1 and CX2 in the circumferential direction are set in a range of several mm or less.

The same applies to each of the even-numbered drawing units U2, U4, and U6, and the circumferential direction is the range of the circumferential positions CX3 to CX4 such that the reference pattern FMa sufficiently crosses the even-numbered drawing lines SL2, SL4, and SL6. While moving to, the light and dark image data is sampled by the reflected light monitor systems of the even-numbered drawing units U2, U4, and U6. However, at the circumferential positions CX3 and CX4, in the encoder system using the encoder head EHa3 or EHb3 (see FIGS. 7 and 8), the digital counter in the counter circuit unit (referred to as 200C) similar to the counter circuit unit 200B is used. It is specified by the movement amount (circumferential length measured under the unit movement amount ΔLxa) after being zero-reset by the origin pattern ZZo.

The range of acquisition of bright and dark image data by the reflected light monitor system is the length (for example, 52 mm) x distance (several mm determined by CX2-CX1 or CX4-CX3) of each of the drawing lines SL1 to SL6 in the Y direction. It is desirable that it is the entire defined two-dimensional area. However, since the amount of data (image storage memory capacity) becomes enormous, it may be limited to a partial area in the Y direction of each of the drawing lines SL1 to SL6 (for example, several mm in the Y direction). For example, regarding the drawing line SL1, a region of several mm including the streak pattern Fxs1 arranged near the start position of the main scan by the spot light and the streak pattern Fxe1 arranged near the end position of the main scan by the spot light. Bright and dark image data may be acquired only for each of the region of several mm including the region and the region of several mm including the linear pattern Fxc1 arranged near the intermediate position of the main scan by the spot light.

FIG. 17 shows, as an example, how the spot light SP that generates the drawing line SL1 relatively two-dimensionally scans the region including the linear pattern Fxc1 in the reference pattern FMa on the outer peripheral surface DRs of the rotating drum DR. .. The oscillation frequency FPL of the beam LB from the light source device LS shown in FIG. 1 (same as the clock signal LTC described in FIG. 11) is 400 MHz (period 2.5 nS), and the effective diameter of the spot light SP by one pulse. ^{When φp (for example, a diameter that becomes 1 / e 2} of the peak intensity in an approximate Gaussian distribution) is 2.5 μm, the spot light SP pulsed every 2.5 nS is in the main scanning direction (Y direction). ), The rotation speed of the polygon mirror PM is set so as to move by the unit scanning movement amount ΔYsp (1.25 μm, which is 1/2 of the diameter φp). Further, the main scanning of the spot light SP along the drawing line SL1 is performed every time the outer peripheral surface of the rotating drum DR moves in the sub-scanning direction (circumferential direction) by a unit feed amount ΔXsp, and the unit feed amount ΔXsp is determined by the unit feed amount ΔXsp. The rotation speed of the rotating drum DR is set so as to be 1.25 μm, which is 1/2 of the effective diameter φp of the spot light SP. In FIG. 17, the sequence of spot light SPs that emit pulses at the same position in the main scanning direction with respect to the sub-scanning direction is defined as the sub-scanning line SL1'.

When the intensity level of the photoelectric signal from the reflected light monitor system (photoelectric sensor DTR) is digitally sampled by the ADC unit 202A (FIG. 11) for each clock pulse of the clock signal LTC, for example, along one drawing line SL1. The signal waveform Wfy as the light / dark image data as shown on the right side of FIG. 17 is acquired. When the drawing line SL1 crosses the linear pattern Fxc1, the regular reflected light due to the irradiation of the spot light SP is lower than the regular reflected light from the surroundings in the linear pattern Fxc1, so that the main scanning direction (Y direction) of the linear pattern Fxc1 ), The level of the signal waveform Wfy changes at each of the edge positions Yac1, Yac2, Yac3, and Yac4. The image processing unit 204A shown in FIG. 11 determines the four edge positions Yac1 to Yac4 by comparing the level of the signal waveform Wfy with an appropriate threshold value Vsz. The position of the spot light SP along the drawing line SL1 is one of the origin signals output from the origin sensor in the drawing unit U1 each time each reflecting surface of the polygon mirror PM becomes an angular position immediately before the start of drawing. The time point at which the pulse is generated is set as the origin position (zero point position), and is specified by the count value of the clock pulse of the clock signal LTC (the address value of the waveform memory that stores the signal waveform Wfy).

The image processing unit 204A calculates the average position PQa1y (= [Yac1 + Yac2 + Yac3 + Yac4] / 4) of the four determined positions Yac1 to Yac4 as the center position in the main scanning direction (Y direction) of the linear pattern Fxc1. .. Further, as shown in the upper part of FIG. 17, the image processing unit 204A identifies the signal waveform Wfx obtained along the sub-scanning line SL1'between the position Yac2 and the position Yac3 in the signal waveform Wfy. The signal waveform Wfx is preferably the sum of the signal waveforms Wfx obtained at each of the plurality of sub-scanning lines SL1'between positions Yac2 and Yac3 in the main scanning direction. The image processing unit 204A determines two edge positions Xac1 and Xac2 in the sub-scanning direction of the linear pattern Fyo by comparing the level of the signal waveform Wfx with an appropriate threshold value Vsz, and further determines the average of the positions Xac1 and Xac2. Position PQa1x (= [Xac1 + Xac2] / 2) is calculated as the center position in the sub-scanning direction of the linear pattern Fyo. The positions Xac1, Xac2, and PQa1x are specified by the measurement information ESa2 and ESb2 (for example, resolution 0.1 μm) measured by the counter circuit unit 200B (FIG. 11) of the encoder system.

By the above arithmetic processing, the coordinate positions (PQa1x, PQa1y) of the center point (intersection) PQa1 (corresponding to the center point Cfm in FIG. 15) of the portion where the straight line pattern Fyo and the streak pattern Fxc1 intersect are determined. Similarly, the ADC unit 202A and the image processing unit 204A determine the coordinate position of the intersection PQs1 between the linear pattern Fxs1 and the linear pattern Fyo located on the left end side (scanning start side) on the drawing line SL1 shown in FIG. , The coordinate position of the intersection point PQe1 between the linear pattern Fxe1 located on the right end side (scanning end side) on the drawing line SL1 and the straight line pattern Fyo is determined. Similarly, for each of the other drawing units U2 to U6, the linear patterns Fxc2 to Fxx6 are obtained by the main scanning of the spot light SP and the rotation (sub-scanning) of the rotating drum DR along each of the drawing lines SL2 to SL6. The coordinate positions of the intersections PQa2 to PQa6 of each of the above and the linear pattern Fyo1 are determined. Further, each coordinate position of the line pattern Fxs2 to Fxs6 located on the scanning start side of each of the drawing lines SL2 to SL6 and the intersection point PQs2 to PQs6 of the linear pattern Fyo, and the scanning end of each of the drawing lines SL2 to SL6. The coordinate positions of the intersection points PQe2 to PQe6 of each of the linear patterns Fxe2 to Fxe6 located on the side and the linear pattern Fyo are determined.

The arrangement states (for example, intervals YJ12, YJ23, YJ34, YJ45, YJ56, etc.) of the linear patterns Fyo, the linear patterns Fxx1 to Fxc6, Fxs1 to Fxs6, and Fxe1 to Fxe6 on the rotating drum DR are shown in FIG. Since it is known as described above, a two-dimensional position error of each of the midpoints CE1 to CE6 of the drawing lines SL1 to SL6 can be obtained based on the measured coordinate positions of the intersections PQa1 to PQa6. Here, in the present embodiment, the reference at the time of pattern drawing is set as the drawing line SL1, and the placement error of the midpoints CE2 to CE6 of the other drawing lines SL1 to SL6 is set with respect to the midpoint CE1. It shall be asked whether it is. Of course, the drawing line SL1 to be used as a reference may be any one of the other drawing lines SL2 to SL6.

FIG. 18 shows a state in which the relative arrangement error with the adjacent drawing line SL2 is determined using the drawing line SL1 in which the spot light SP is scanned between the scanning start position Yss1 and the scanning end position Yse1 as a reference at the time of drawing. It is a figure explaining. For each of the other drawing lines SL3 to SL6, there are cases where the relative arrangement error is determined with reference to the drawing line SL1, the relative arrangement error of the drawing line SL3 with respect to the drawing line SL2, and the drawing line SL3. Determines the relative placement error of the drawing line SL4 based on the drawing line SL4, the relative placement error of the drawing line SL5 based on the drawing line SL4, and the relative placement error of the drawing line SL6 based on the drawing line SL5. There are cases where it is done. With either determination method, the relative position error amount (vector amount) of each of the drawing lines SL1 to SL6 can be specified.

In FIG. 18, it is assumed that the drawing lines SL1 and SL2 have no rotation error and are set without inclination with respect to the Y axis. Further, the intersection PQa1 between the straight line pattern Fyo and the linear pattern Fxc1 and the midpoint CE1 of the drawing line SL1 are actually displaced from each other, but in FIG. 18, since the drawing line SL1 is used as a reference, the intersection PQa1 and It is represented in a state where it coincides with the midpoint CE1. Further, the design distance between the drawing line SL1 (midpoint CE1) and the drawing line SL2 (midpoint CE2) in the sub-scanning direction (circumferential direction) is set to the specified distance ΔLM, and the intersection PQa1 of the midpoint CE1 of the drawing line SL1 is set. The position PQa1x (FIG. 17) in the sub-scanning direction (circumferential direction) is the position CXs measured by the encoder system, and the point measured by the encoder system is the position CXe that advances in the sub-scanning direction by a specified distance ΔLM from the position CXs. And.

FIG. 18 shows a state in which the linear pattern Fyo of the reference pattern FMa is positioned at the position CXe in the sub-scanning direction, and the drawing line SL2 (midpoint CE2) is precise on the linear pattern Fyo due to the relative positional error. It has a position error of Δxx2 on the negative side in the sub-scanning direction with respect to the linear pattern Fyo (position CXe) without overlapping with. The position error Δxx2 is the coordinate position (PQa2x) in the sub-scanning direction of the intersection PQa2 of the midpoint CE2 of the drawn line SL2 and the coordinate position (PQa1x) in the sub-scanning direction of the intersection PQa1 of the midpoint CE1 of the drawing line SL1. It is obtained by the calculation of Δxx2 = (PQa2x-PQa1x) -ΔLM based on the difference length between the above and the specified distance ΔLM.

The design distance between the midpoint CE1 of the drawing line SL1 and the midpoint CE2 of the drawing line SL2 in the main scanning direction (Y direction) is the same as the linear pattern Fxc1 of the reference pattern FMa if there is no relative positional error. The distance between the line pattern Fxc2 and the Y direction coincides with YJ12, and the midpoint CE2 of the drawing line SL2 overlaps the line pattern Fxc2 in the Y direction with reference to the drawing line SL1 (midpoint CE1). However, in reality, as shown in FIG. 18, a positional error Δyy2 relative to the Y direction may occur between the midpoint CE2 of the drawing line SL2 and the streak pattern Fxc2. The position error Δyy2 measures the Y-direction positions of the linear patterns Fxx2, Fxs2, and Fxe2 starting from the scanning start position Yss2 between the scanning start position Yss2 and the scanning end position Yse2 of the drawing line SL2. Is required by.

As described above, the relative two-dimensional position error (Δxx2, Δyy2) of the midpoint CE2 of the drawing line SL2 with reference to the midpoint CE1 of the drawing line SL1 is determined. Therefore, the position error of the midpoint CE2 of the drawing line SL2 is set to ΔFS2 (Δxx2, Δyy2), and the position error of each of the midpoints CE3 to 6 of the drawing lines SL3 to 6 based on the midpoint CE1 of the drawing line SL1. Let be ΔFS3 (Δxx3, Δyy3), ΔFS4 (Δxx4, Δyy4), ΔFS5 (Δxx5, Δyy5), and ΔFS6 (Δxx6, Δyy6). Each of these position errors ΔFS2 to ΔFS6 can be set to a dozen μm or less by adjustment at the time of assembling the apparatus. When drawing a pattern by each of the drawing lines SL1 to SL6, the timing of pattern drawing by each of the drawing units U2 to U6 and the position of the drawing lines SL1 to SL6 in the sub-scanning direction are finely adjusted based on their position errors ΔFS2 to ΔFS6. By doing so, the splicing error between the patterns drawn by each of the drawing lines SL1 to SL6 is less than a fraction of the minimum drawable line width, for example, when the minimum line width is 5 μm, the splicing error amount is 1 μm. It can be less than (± 0.5 μm at 3σ).

Further, in FIG. 18, the drawing lines SL1 and SL2 (and SL3 to SL6) are not tilted with respect to the Y axis (tilt error is zero), but the presence / absence of tilt and the amount of tilt error are determined by the drawing line SL1. Is obtained by measuring the positional error in the sub-scanning direction (circumferential direction) between the intersection position of the linear pattern Fyo and the linear pattern Fxs1 and the intersection position of the linear pattern Fyo and the linear pattern Fxe1. For the other drawing lines SL2 to SL6, the presence / absence and amount of tilt error can be determined by the same measurement. The inclination error of each drawing line SL1 to SL6 is a rotation in which each of the drawing units U1 to U6 shown in FIG. 3 is slightly rotated around the line segments LE1 to LE6 in FIG. 3 passing through each of the midpoints CE1 to CE6. It is corrected by the drive mechanism. In the initial state, each of the drawing lines SL2 to SL6 is set to be parallel to the Y-axis within an allowable error range.

FIG. 19 is a diagram schematically exaggerating the calibration information (arrangement error, etc.) determined or set by the above step 304. FIG. 19 shows the installation errors ΔC1 to ΔC4 (vector) of the center points CC1 to CC4 of each imaging region DIS'of the alignment systems ALG1 to ALG4 determined with reference to the reference marks RM1 to RM4 of the reference bar member RB. It is shown. Further, in FIG. 19, when the drawing line SL1 (the drawing line in which drawing is first performed with respect to the transport direction of the substrate P) among the six drawing lines SL1 to SL6 is used as a reference, the alignment system ALGn described with reference to FIG. The position CXA in the sub-scanning direction (circumferential direction) with respect to each center point CCn and the position CXs in the sub-scanning direction (circumferential direction) with respect to the drawing line SL1 (midpoint CE1) described with reference to FIG. The interval is called the reference length (baseline length) ΔBSL. The reference length ΔBSL is determined by changes in the device environment (temperature and atmospheric pressure), thermal deformation of mechanical parts due to the influence of heat sources (actuators that drive the rotating motor of polygon mirror PM, optical members, etc.) in the drawing units U1 to U6, etc. May fluctuate on the order of microns.

As described above, the coordinate positions (or placement errors) of the center points CC1 to CC4 of the alignment systems ALG1 to ALG4 on the outer peripheral surface of the rotating drum DR are determined by the reference pattern Fxa (or the arrangement error) on the rotating drum DR. It is obtained based on the intersection of each of the linear patterns Fxe1, Fxe3, Fxe4, and Fxe6 of Fxb to Fxh) and the linear pattern Fyo. Further, the coordinate positions of the midpoints CE1 to CE6 of the drawing lines SL1 to SL6 on the outer peripheral surface of the rotating drum DR are based on the midpoint CE1 of the drawing line SL1 as described in FIG. Positional errors ΔFS2 (Δxx2, Δyy2), ΔFS3 (Δxx3, Δyy3), ΔFS4 (Δxx4, Δyy4), ΔFS5 (Δxx5, Δyy5), ΔFS6 (Δxx6) of each midpoint CE2 to 6 of the obtained drawing lines SL2 to SL6. , Δyy6).

From the above, with reference to the reference pattern Fxa (or Fxb to Fxh), the center points CC1 to CC4 of each of the alignment systems ALG1 to ALG4 and the midpoints CE1 to CE6 of each of the drawing lines SL1 to SL6 Relative positional relationships are identified. Further, the two-dimensional relative positional relationship between the reference pattern Fxa (or Fxb to Fxh) on the rotating drum DR and the reference marks RM1 to RM4 on the reference bar member RB is also an image with each of the alignment systems ALG1 to ALG4. It is obtained with high accuracy based on the image analysis result by the analysis unit 206 and the like. Therefore, by executing the sequence from steps 300 to 304, the coordinate positions of the drawing lines SL1 to SL6 (particularly the coordinate positions of the midpoints CE1 to CE6) with reference to the reference marks RM1 to RM4 on the reference bar member RB. ), Each placement error (positional deviation error) is indirectly specified.

This means that the position of the pattern drawn by each of the drawing lines SL1 to SL6 at the time of pattern drawing, particularly the position of the pattern for 1st exposure on the substrate P, is precisely aligned with the arrangement of the reference marks RM1 to RM4 on the reference bar member RB. It is possible to correct the drawing timing of each of the drawing units U1 to U6 and fine-tune the drawing lines SL1 to SL6 in the sub-scanning direction (correction of the inclination amount of the parallel flat plate HVP shown in FIG. 3) so as to imitate. Means. The measurement control unit 210 shown in FIG. 11 stores various calibration information (arrangement errors ΔC1 to ΔC4, position errors ΔFS2 to ΔFS6, reference length ΔBSL, specified distance ΔLM, etc.) obtained by executing steps 300 to 304. .. In FIG. 19, the reference length ΔBSL is the average position in the sub-scanning direction of the odd-numbered drawing lines SL1 (midpoint CE1), SL3 (midpoint CE3), and SL5 (midpoint CE5), and the even-numbered drawing lines. It may be the distance between the intermediate position CXs'and the position CXA with the average position in the sub-scanning direction of the drawing lines SL2 (midpoint CE2), SL4 (midpoint CE4), and SL6 (midpoint CE6).

[Step 306]
Next, in step 306 of the operation sequence shown in FIG. 12, the paper passing operation of the substrate P is performed. In the paper passing work, a long substrate P (with a photosensitive layer formed on the surface) wound on a supply roll mounted on a roll-to-roll processing device is passed along a transport path in the processing device. After that, the tip of the substrate P is wound around the recovery roll, and the substrate P is set up so that it can be conveyed with a predetermined tension without meandering.

[Step 308]
Next, in step 308 of the operation sequence shown in FIG. 12, it is determined whether the passed substrate P is for the first (1st) exposure or the second (2nd) exposure, and in the case of the substrate P for the 1st exposure. The process proceeds to step 310, and in the case of the substrate P for 2nd exposure, the process proceeds to step 316. The 1st exposure is the first exposure on the photosensitive layer of the substrate P on which no pattern is formed in the exposed region DPA as shown in FIG. 13 and no alignment marks MK1 to MK4 are formed. It means exposing the pattern for the layer. In the 2nd exposure, a new pattern to be superimposed on the background pattern is exposed on the photosensitive layer of the substrate P on which some background pattern is formed in the exposed area DPA and the alignment marks MK1 to MK4 are formed. Means to do.

[Step 310]
When the 1st exposure is determined in step 308, in step 310 of the operation sequence shown in FIG. 12, each of the drawing units Un (n = 1 to 6) forms the pattern for the first layer and the alignment marks MK1 to MK4. , Setup such as downloading drawing data and adjusting the intensity of the beam from the light source device LS is executed so as to draw along the drawing lines SL1 to SL6. Further, at the time of the 1st exposure, since the reference alignment marks MK1 to MK4 are not formed on the substrate P, the pattern drawing is performed based on the positions of the drawing lines SL1 to SL6. In the present embodiment, the pattern drawn on the substrate P by each of the drawing lines SL1 to SL6 based on various calibration information stored in the measurement control unit 210 is the reference mark RM1 on the reference bar member RB. It can be set up so that it is arranged with reference to ~ RM4. Note that [210] added next to step 310 in FIG. 12 means to use various calibration information stored in the measurement control unit 210.

[Step 312]
When the setup for the 1st exposure is completed in step 310 of FIG. 12, the rotary drum DR is rotationally driven so that the substrate P moves in the sub-scanning direction at a set speed. Similar to steps 300 to 304, [EH] added next to step 312 in FIG. 12 uses the measurement position information by the encoder measurement system (here, mainly the encoder heads EHa2, EHa3, EHb2, EHb3). Means. Further, [LS] added next to step 312 is a beam whose intensity is modulated according to the drawing data corresponding to the pattern for 1st exposure from the light source device LS shown in FIG. 1, respectively, in the drawing units U1 to U6. It means that it is supplied in a time-division manner toward. By the 1st exposure in step 312, the photosensitive layer on the substrate P is sequentially exposed to the pattern for the first layer of the electronic device on each of the plurality of exposed region DPAs in the arrangement as shown in FIG. , Alignment marks MK1 to MK4 are exposed.

At the time of this 1st exposure, the two-dimensional positional relationship of the patterns drawn on each of the drawing lines SL1 to SL6 is based on the reference bar member RB (reference marks RM1 to RM4) as a result of the above calibration. Can be set precisely. That is, the drawing position of the pattern drawn on each of the drawing lines SL1 to SL6 on the substrate P is precisely copied to the arrangement of the reference marks RM1 to RM4 of the stable reference bar member RB which is an absolute reference. Can be placed. This also applies to each of the alignment marks MK1 to MK4 drawn on the substrate P by the 1st exposure, and the absolute position and relative positional relationship of each of the alignment marks MK1 to MK4 on the substrate P are used as a reference. It precisely follows the arrangement of the reference marks RM1 to RM4 of the bar member RB.

[Step 314]
If an error or error occurs during the operation of the 1st exposure of the substrate P to the photosensitive layer, the main control unit (main computer) of the drawing device sequentially outputs log information that can identify the error or the occurrence status or state of the error. collect. After the exposure of the substrate P from the supply roll is completed, the main control unit analyzes the collected log information, and if it is determined that the initial calibration state needs to be readjusted, the substrate of the next supply roll is used. Before passing the paper, one of

steps

300 and 302 of the calibration operation is executed again. Even if it is determined that readjustment is not necessary as a result of the analysis of the log information, it is preferable to perform the calibration operation in step 304 for the exposure of the substrate of the next supply roll. This is because even if the exposure of the next supply roll to the substrate is the 1st exposure, the type of electronic device manufactured on the substrate may change, or the exposure of the next supply roll to the substrate may be the 2nd exposure. ..

[Step 316]
In step 308 above, when the pattern to be exposed to the photosensitive layer of the substrate P is for 2nd exposure, in step 316 of the operation sequence shown in FIG. 12, each of the drawing units Un (n = 1 to 6) is the first. Setup such as downloading drawing data and adjusting the intensity of the beam from the light source device LS is executed so that the patterns of the second and subsequent layers are drawn along the drawing lines SL1 to SL6. [210] added next to step 316 means to use various calibration information stored in the measurement control unit 210. Further, since the substrate P for 2nd exposure is subjected to a process such as a wet treatment or a heat treatment, it may be accompanied by a steady expansion / contraction error or deformation error that can be predicted. Therefore, in step 316, if necessary, it is possible to suppress deterioration of splicing exposure accuracy, superimposition accuracy, etc. by each of the drawing units U1 to U6 based on steady prediction values regarding expansion / contraction error and deformation error of the substrate P. Information for correcting the drawing timing is also set up. However, correction of drawing timing during the actual drawing operation of the 2nd exposure, correction of each minute rotation of the drawing units U1 to U6, correction of each minute shift of the drawing lines SL1 to SL6 by the parallel flat plate HVP in FIG. 3, or drawing. The correction of the magnification and the like are determined based on the position measurement results of the alignment marks MK1 to MK4 on the substrate P.

[Step 318] and [Step 320]
In the present embodiment, as shown in FIGS. 4 and 7, when each position of the alignment marks MK1 to MK4 on the substrate P is detected by the alignment systems ALG1 to ALG4, each of the drawing lines SL1 to SL6 is immediately detected. Pattern drawing is performed by. In particular, since a plurality of alignment marks MK1 and MK4 formed on both ends in the width direction (Y direction) of the substrate P are provided at regular intervals in the sub-scanning direction, the mark detection operation by the alignment systems ALG1 and ALG4 and the mark detection operation can be performed. The pattern drawing operation by the drawing units U1 to U6 is superimposed. [ALGn] added next to step 318 in FIG. 12 utilizes the measurement position information of each of the alignment marks MKn (n = 1 to 4) detected by the alignment system ALGn (n = 1 to 4). Meaning, [EH] means to use the measurement position information by the encoder measurement system (encoder heads EHa1 to EHa3, EHb1 to EHb3).

Further, [RB] added next to step 318 means that the reference mark RNn (n = 1 to 4) of the reference bar member RB is used when measuring the position of the alignment mark MKn, but it is not always used. do not have to. The alignment system ALGn has a reference center point CCn in the imaging region DIS'. Therefore, normally, as shown in FIG. 9C, the alignment measurement is completed by measuring the two-dimensional displacement amount between the enlarged image MKn'of the alignment mark MKn and the center point CCn. However, during the exposure operation over a long period of time, the position of the alignment system ALGn drifts, and the characteristics of the internal optical member fluctuate due to changes in temperature and atmospheric pressure, so that the center specified at the time of calibration Each of the points CCn may change position relatively. Even if the amount of the position change is on the order of microns, it becomes the amount of error of the measurement position of the alignment mark MKn as it is, so that the splicing error and the superposition error are deteriorated.

In order to avoid such a problem, in the alignment measurement in step 318, when the alignment mark MKn appears in each imaging region DIS'of the alignment system ALGn, it is set at a substantially fixed position in the imaging region DIS'. An enlarged image RNn'of the reference mark RNn and an enlarged image MKn'of the alignment mark MKn are sampled at the same time, and each of the alignment mark MKn is based on the center point CRn (n = 1 to 4) of the reference mark RNn. The two-dimensional misalignment error of the center point of the magnified image MKn'is measured. As shown in FIG. 13, four alignment marks MK1 to MK4 are formed on each of the margins of the front end portion and the rear end portion of the exposed region DPA on the substrate P in the transport direction by the patterning of the 1st exposure. Along a line extending in the width direction of the substrate P, a position corresponding to the arrangement of each imaging region DIS'in each of the detection regions AD1 to AD4 (see FIG. 7) of the four alignment systems ALG1 to ALG4, that is, the reference bar. It is formed at a position corresponding to each arrangement of the reference marks RM1 to RM4 of the member RB.

Therefore, immediately before (or immediately after) the 2nd exposure to each exposed area DPA, the relative positional deviation between the four reference marks RM1 to RM4 and the four alignment marks MK1 to MK4 imaged at the same timing is obtained, and the relative positional deviation is obtained. By obtaining the positional deviations of the alignment marks MK1 and MK4 on both sides of the exposed area DPA, the expansion and contraction error and the deformation error in the Y direction of the exposed area DPA are sequentially estimated, and the drawing lines SL1 to SL6 are adjusted accordingly. The drawing position of each pattern is sequentially fine-tuned during the 2nd exposure in step 320. In the fine adjustment, the drawing timing is corrected by the spot light SP, the shift correction of the drawing line SLn is performed by the parallel flat plate HVP, the minute rotation correction of the drawing unit Un, and the drawing magnification correction are performed. In addition, [EH] added next to step 320 in FIG. 12 means to use the measurement position information by the encoder measurement system (encoder heads EHa1 to EHa3, EHb1 to EHb3), and further added next to step 320. [LS] supplies a beam whose intensity is modulated according to the drawing data corresponding to the pattern for 2nd exposure from the light source device LS shown in FIG. 1 to each of the drawing units U1 to U6 in a time-divided manner. Means that.

In addition, steps 318 and 320 of FIG. 12 represent the 2nd exposure to one exposed area DPA on the substrate P, and whether or not the 2nd exposure to a plurality of exposed area DPA to be exposed on the substrate P is completed is completed. , It is determined in the next step 322. When it is determined in step 322 that the 2nd exposure is continuously performed on the next exposed area DPA, steps 318 and 320 are repeatedly executed.

[Step 324]
When the 2nd exposure to all the exposed area DPA is completed, in step 324, various log information (error information, error information, etc.) during the exposure operation is collected, and a recovery roll or the like around which the 2nd exposed substrate P is wound. Is removed from the drawing device and the substrate P is transported to the next process. When the substrate P is removed from the drawing apparatus, the calibration operations of

steps

302 and 304 above are executed again. However, when the length of the 2nd exposed substrate P in the long direction is short and the operation duration of the 2nd exposure is relatively short, the operation from step 304 or step 306 is executed after step 324. You can also do it. On the contrary, when the continuous operation of the 2nd exposure lasts for a long time, the calibration operation from step 300 may be executed.

In the alignment measurement in step 318, a part of the alignment mark MKn on the substrate P is damaged due to the influence of a process or the like, or foreign matter (dust) having the same size as the line width dimension of the mark is attached in the vicinity. As a result, the alignment system ALGn may not recognize the image well, and a detection error may occur. When such a detection error continues over a predetermined transport distance of the substrate P, the operation of the 2nd exposure (transportation of the substrate P in the forward direction by the rotating drum DR) is temporarily stopped, and the retry operation is performed. You can also. In the retry operation, the image processing conditions (parameters) of the magnified image MKn'of the alignment mark MKn by the image analysis unit 206 shown in FIG. 11 and the adjustment and wavelength of the intensity of the illumination light ILb from the illumination system ILU shown in FIG. 9 are adjusted. After shifting the band and the like, the substrate P is returned in the reverse direction by a certain distance, and then the alignment measurement and the 2nd exposure are performed again while transporting in the forward direction. An example of such a retry operation is disclosed in, for example, International Publication No. 2018/030357.

As described above, among the exposure sequences according to the present embodiment shown in FIG. 12, particularly in the 1st exposure sequence (steps 310 and 312), the absolute positions of the drawing lines SL1 to SL6 by each of the drawing units U1 to U6 And the relative positional relationship is calibrated with reference to the reference marks RM1 to RM4 of the reference bar member RB that are precisely positioned with respect to the central axis AXo (and the outer peripheral surface DRs) of the rotating drum DR as a transport system. Therefore, it is possible to reduce that the shape (rectangle) of the entire exposed region DPA drawn on the sheet-shaped substrate P is deformed into a parallel quadrilateral shape, a saddle shape, or an arch shape.

Further, in the 2nd exposure sequence (

steps

316, 318, 320) of the exposure sequence according to the present embodiment, the 2nd exposure sequence is not affected by the drift of the relative positional relationship of the detection regions AD1 to AD4 of the alignment systems ALG1 to ALG4. Even during exposure, each position of the alignment marks MK1 to MK4 on the substrate P can be detected with reference to the reference marks RM1 to RM4 of the reference bar member RB. Therefore, even if the 2nd exposure is continuously performed for a long time and the alignment system ALGn may drift due to temperature change, atmospheric pressure change, etc., the first of the plurality of exposed region DPAs lined up on the substrate P From the exposed region DPA to the last exposed region DPA, the position detection accuracy of the substrate P and the superposition accuracy of the 2nd exposure can be maintained constant.

[Modification 1]
FIG. 20 shows a modified example of the arrangement of the alignment system ALGn of the drawing apparatus shown in FIG. 7, and shows the state when the alignment system ALGn is increased from 4 to 7, with the XY plane of the orthogonal coordinate system XYZ. It is a figure seen in a parallel plane. In FIG. 20, members and structures having the same functions as those in FIG. 7 are designated by the same reference numerals. In FIG. 20, each of the detection regions AD1 to AD7 of the seven alignment systems ALG1 to ALG7 are arranged at predetermined intervals in the Y direction. Therefore, in this modification, the reference mark (reference index mark) RMn formed on the reference surface RBa of the reference bar member (reference index member) RB is also the distance between the design detection regions AD1 to AD7 in the Y direction. The reference marks RM1 to RM7 are formed at each of the seven positions corresponding to.

Similar to FIG. 7, when the maximum dimension in the Y direction that can be continuously exposed by the six drawing lines SL1 to SL6 is WAY, the detection area AD1 located on the negative side of the seven detection areas AD1 to AD7 in the Y direction. Is within the maximum dimension WAY and is located near the end of scanning of the drawing line SL1, and the detection area AD7 located on the positive side in the Y direction is within the maximum dimension WAY and is near the end of scanning of the drawing line SL6. To position. Further, regarding the position in the Y direction, the detection area AD2 is arranged at the joint exposure portion formed by the scanning start point of the drawing line SL1 and the scanning start point of the drawing line SL2, and the detection area AD3 is arranged as the scanning end point of the drawing line SL2. The detection area AD4 is arranged in the joint exposure portion formed by the scanning end point of the drawing line SL3 and the scanning start point of the drawing line SL3, and the detection area AD5 is arranged in the joint exposure portion formed by the scanning start point of the drawing line SL3 and the scanning start point of the drawing line SL4. The detection area AD6 is arranged in the joint exposure portion formed by the scanning end point of the drawing line SL4 and the scanning end point of the drawing line SL5, and the detection area AD6 is located in the joint exposure portion formed by the scanning start point of the drawing line SL5 and the scanning start point of the drawing line SL6. Be placed.

As for the alignment marks MKn formed on the substrate P, seven alignment marks MK1 to MK7 are arranged corresponding to the respective arrangements of the detection regions AD1 to AD7 in the width direction (Y direction) of the substrate P. Of course, a large number of alignment marks MK1 and MK7 formed on both sides of the substrate P in the width direction are formed in a row at regular intervals (for example, 10 mm) along the long direction (sub-scanning direction) of the substrate P. Has been done. Seven alignment marks MK1 to MK7 arranged in a row in the Y direction can be formed in the margin between the exposed region DPA and the exposed region DPA on the substrate P in the long direction, and immediately before the 2nd exposure. It is possible to estimate the deformation state of the exposed area DPA in detail. In the case of 2nd exposure to the substrate P, which is less deformed by wet treatment or heat treatment, the alignment marks MK2 to MK6 are used for some of the alignment systems ALG2 to ALG6 during the alignment measurement in step 318 of FIG. It is also possible to omit (skip) the detection of.

[Modification 2]
FIG. 21 shows deformation of the arrangement relationship of the drawing lines SL1 to SL6 by the drawing units U1 to U6 shown in FIGS. 7, 16 and 20 on the substrate P (or the outer peripheral surfaces DRs of the rotating drum DR). An example is shown. In this modification, the drawing units U1 to U6 are arranged so that the joint portions of the patterns drawn on the substrate P by each of the drawing lines SL1 to SL6 overlap (overlap) by a certain dimension in the Y direction. ing. Since each of the drawing units U1 to U6 has the same configuration, the length (drawing length) ΔMLs at which the pattern can be drawn by each of the drawing lines SL1 to SL6 is the same. Within the drawing length ΔMLs, each of the range of a constant length ΔOLs from the scanning start point of the spot light SP or a constant length ΔOLs from the scanning end point of the spot light SP is covered by the overlapping regions OL12, OL23, OL34, OL45, OL56 (OL45 and OL56 are not shown).

The overlap region OL12 is a range in which the length ΔOLs on the scanning start point side of the spot light SP along the drawing line SL1 and the length ΔOLs on the scanning start point side of the spotlight SP along the drawing line SL2 overlap. be. Similarly, in the overlap region OL23, the length ΔOLs on the scanning end point side of the spot light SP along the drawing line SL2 and the length ΔOLs on the scanning end point side of the spot light SP along the drawing line SL3 overlap. The overlap region OL34 includes the length ΔOLs on the scanning start point side of the spot light SP along the drawing line SL3 and the length ΔOLs on the scanning start point side of the spot light SP along the drawing line SL4. Is the overlapping range. Although not shown in FIG. 21, the other overlap region OL45 includes the length ΔOLs on the scanning end point side of the spot light SP along the drawing line SL4 and the scanning end point of the spot light SP along the drawing line SL5. The overlap region OL56 is a range in which the length ΔOLs on the side overlaps, and the overlap region OL56 is the scanning of the length ΔOLs on the scanning start point side of the spot light SP along the drawing line SL5 and the spot light SP along the drawing line SL6. This is the range where the length ΔOLs on the start point side overlaps.

The length ΔOLs of the overlap region can be about 0.5 to 2% of the drawing length ΔMLs of the drawing line. As an example, when the drawing length ΔMLs is 50.0 mm, the length ΔOLs is about 0.25 mm to 1.00 mm. Further, the patterns drawn in the overlap areas OL12, OL23, OL34, OL45, and OL56 are exposed so as to have the same shape and precisely overlap at the same position on the substrate P, so that the pattern is doubled when drawn as it is. It will be the amount of exposure. Therefore, the drawing data is modified so that the pattern included in each overlapping region has a checkered flag shape (checkerboard pattern), for example, as shown in FIG. 22.

FIG. 22 shows a state of the same pattern (pixel array on two-dimensional drawing data) that is overexposed in the overlap region OL12 between the scanning start point side of the drawing line SL1 and the scanning start point side of the drawing line SL2. show. The dimensions of the pixel PIX on the 1-bit substrate P of the drawing data are set according to the effective diameter of the spot light SP scanned along the drawing lines SL1 and SL2. The drawing data is defined by a set of the smallest square pixel PIX that can be drawn, and whether the spot light SP (pulse shape) is irradiated or not irradiated to the pixel PIX is determined by the 1-bit logical value ". It is represented by "0" or "1". As shown in FIG. 22, in the drawing data corresponding to the pattern existing in the overlap region OL12, the pixel PIX is the pixel PIXa (white square) having a logical value “0” (non-irradiation) and the logical value “1”. It is decomposed into a checkered pattern by the pixels PIXb (diagonal squares) of (irradiation). Further, the pattern in the overlap area OL12 drawn by the drawing line SL1 and the pattern in the overlap area OL12 drawn by the drawing line SL2 are complementary to the pixels PIXa and the pixels PIXb arranged in a checkered pattern. It is set in a (complementary) relationship.

In this way, the drawing data of the pattern existing in the overlap region OL12 (OL23, OL34, OL45, OL56) is decomposed into a complementary relationship with the logical value for each pixel PIX, so that the overlap on the substrate P is performed. The pattern exposed to the region is less likely to have a conspicuous stepped shape or a conspicuous thickening of the line width due to the influence of a slight splicing error.

[Second Embodiment]
FIG. 23 is a diagram showing an optical configuration of the alignment system ALGn according to the second embodiment, and the Cartesian coordinate system XYZ is set in the same manner as in FIG. 2 in the first embodiment. Further, in FIG. 23, the same members and configurations as those in the first embodiment are designated by the same reference numerals. In the present embodiment, as shown in FIGS. 4 to 6 in the first embodiment, the reference bar member RB is not provided on the object surface side (board P side) of the objective lens system OBL. An intermediate image plane was formed in the optical path of the alignment system ALGn, and the intermediate image plane was arranged at a position corresponding to the intermediate image plane.

The alignment system ALGn shown in FIG. 23 includes a plane mirror Mb arranged from the substrate P side, a first imaging optical system GLo, a cube-shaped first beam splitter BS1 (synthetic optical member), and a second imaging optical system Gd. It is composed of a second beam splitter BS2. Illumination light ILb (light in the non-photosensitive wavelength range) from the illumination system ILU (see FIGS. 6 and 9) (not shown) is reflected by the second beam splitter BS2 and travels coaxially with the optical axis AXs. 2 It is incident on the imaging optical system Gd and is divided into a component transmitted through the first beam splitter BS1 and a component reflected. The illumination light ILb transmitted through the first beam splitter BS1 enters the first imaging optical system GLo through the intermediate image plane Pss, is reflected by the plane mirror Mb, and is reflected on the surface of the substrate P (or the outer periphery of the rotating drum DR). The detection region ADn on the surface DRs) is illuminated with a uniform illuminance.

When the alignment mark MKn of the substrate P (or the reference patterns FMa to FMh on the outer peripheral surface DRs of the rotating drum DR) appears in the detection region ADn, the reflected light from the alignment mark MKn (or the reference pattern FMa to FMh) appears. Is incident on the first imaging optical system GLo, and the first imaging optical system GLo forms an image MKn'(or an image of the reference patterns FMa to FMh) of the alignment mark MKn on the intermediate image plane Pss. The first imaging optical system GLo has a low magnification (for example, a magnification of 1 to 2 times or a reduction of 0.75 times) so that the working distance (working distance) on the substrate P side can be relatively large. Is set to. The reflected light from the alignment marks MKn (or reference patterns FMa to FMh) imaged on the intermediate image plane Pss passes through the first beam splitter BS1 and is incident on the second imaging optical system Gd.

The reflected light from the alignment mark MKn (or reference patterns FMa to FMh) that has passed through the second imaging optical system Gd passes through the second beam splitter BS2 to become an imaging luminous flux Bma, and the imaging is omitted. It reaches the imaging surface of the element DIS. The second imaging optical system Gd has a conjugate relationship (imaging relationship) between the intermediate image plane Pss and the image pickup surface of the image sensor DIS, and an enlarged image of the intermediate image formed on the intermediate image plane Pss is captured by the image sensor DIS. Reimage on the image pickup surface of.

Further, the illumination light ILb reflected by the first beam splitter BS1 through the second imaging optical system Gd illuminates the reference mark RMn formed on the reference surface RBa of the reference bar member RB with a uniform illuminance distribution. The reference surface RBa of the reference bar member RB is arranged so as to coincide with the surface Pss' optically corresponding to the intermediate image surface Pss with the first beam splitter BS1 interposed therebetween. Therefore, the reference surface RBa (plane Pss') has a conjugate relationship (imaging relationship) with the image pickup surface of the image pickup device DIS via the second imaging optical system Gd. Therefore, the reflected light from the reference mark RMn illuminated by the illumination light ILb reaches the imaging surface of the imaging element DIS as an imaging luminous flux Bma via the first beam splitter BS1 and the second imaging optical system Gd. Then, the enlarged image of the reference mark RMn and the enlarged image of the alignment mark MKn (or the reference patterns FMa to FMh) can be simultaneously imaged on the imaging surface. Also in the case of the alignment system ALGn in FIG. 23, as described in FIG. 9 above, by exchanging the illumination field diaphragm FAn in the illumination system ILU that emits the illumination light ILb, the enlarged image of the reference mark RMn and the alignment mark It is possible to switch between simultaneous imaging of both the magnified image of MKn (or reference patterns FMa to FMh) and selective imaging of only one of them.

As described above, also in the present embodiment, since a plurality of alignment system ALGn are arranged in the Y direction, the reference bar member RB crosses the space below the first beam splitter BS1 of each of the plurality of alignment system ALGn in the Y direction. It will be extended as. Further, as shown in FIG. 4 of the first embodiment, the reference bar member RB in the present embodiment does not have to be arranged near the outer peripheral surfaces DRs (board P) of the rotating drum DR. , The external dimensions of the reference bar member RB in the XZ plane can be increased to increase the rigidity, and the support mechanism portion for fixing the reference bar member RB (support frame portion 100 and support plate portion in FIG. 8). 102A, 102B, 103A, 103B, corresponding to the connecting

bar members

104a, 104b, 104c) can be enlarged to increase the rigidity.

[Third Embodiment]
FIG. 24 is a diagram showing an optical configuration of the alignment system ALGn according to the third embodiment, and the orthogonal coordinate system XYZ is set in the same manner as in FIGS. 2 and 23 above. Further, in FIG. 24, the same members and configurations as those in the first embodiment and the second embodiment are designated by the same reference numerals. In the present embodiment, an optical microscope having a working distance (working distance) of 10 cm or more is used as the alignment system ALGn. Such a microscope is sold as a machine vision lens by Moritex Corporation, for example, and it can also be used.

In the present embodiment, as shown in FIG. 24, the entire alignment system ALGn is fixed to the support bracket 400 made of metal or ceramics having a low coefficient of thermal expansion. The support bracket 400 is formed in a plate shape parallel to the XZ plane, and is fixed to a structural portion (metrology frame) connected to the support frame portion 100 in FIG. The alignment system ALGn includes a plane mirror Mb arranged to face the substrate P (the outer peripheral surface of the rotating drum DR), a holding hardware 401 for fixing the plane mirror Mb to the support bracket 400, and an objective lens system OBL of the alignment system ALGn. A plate-type (parallel plate-like) beam made of a transmissive optical glass material such as quartz, which is arranged at an angle θe (θe> 0) with respect to a plane perpendicular to the optical axis AXs in the optical path between the plane mirror Mb. Splitter BS1 (synthetic optical member), holding hardware 402 fixing the beam splitter BS1 to the support bracket 400, optical fiber bundle 404 that guides the illumination light ILb from the illumination system ILU so as to epiilluminate the detection region ADn of the alignment system ALGn. The illumination light ILb from the tip end (ejection end) 404a of the optical fiber bundle 404 is reflected toward the objective lens system OBL, and is reflected by the substrate P or the like and incident through the objective lens system OBL. It is composed of a beam splitter BS2, a lens system Gb, and an image pickup element DIS.

In the present embodiment, about 10 to 30% of the intensity of the illumination light ILb emitted from the objective lens system OBL depends on the angle θe and is the surface (optical splitting surface) of the plate-type beam splitter BS1. , Photosynthetic surface) Reflected by Bsp and heads toward the reference bar member RB. The reference surface RBa of the reference bar member RB is optically set at a position corresponding to the surface of the substrate P, and the detection region ARn set on the reference surface RBa is illuminated with a uniform illuminance distribution by a part of the illumination light ILb. NS. The reflected luminous flux Bmr generated at the reference mark RMn arranged in the detection region ARn reaches the beam splitter BS1 along the optical axis AXs', is reflected by the surface Bsp, and becomes the imaging luminous flux Bma, which becomes the objective lens system OBL. Incident to. Also in this embodiment, the image pickup surface (imaging area DIS') of the image pickup device DIS has a conjugate relationship (imaging relationship) with the surface of the substrate P, and also has a conjugate relationship (imaging relationship) with the reference surface RBa of the reference bar member RB. Relationship) is set. In the configuration of FIG. 24, the plate-type beam splitter BS1 is a non-polarizing type, and a glass material other than quartz may be used.

Therefore, when the illumination light ILb is projected from the emission end 404a of the optical fiber bundle 404, both the detection region ADn set on the substrate P and the detection region ARn set on the reference bar member RB are simultaneously illuminated by the illumination light ILb. Illuminated by epi-illumination. Therefore, in the image pickup region DIS'of the image pickup element DIS, an image of the alignment mark MKn of the substrate P appearing in the detection region ADn, or an image of the reference patterns FMa to FMh on the rotating drum DR, and the image of the reference patterns FMa to FMh in the detection region ARn. The image of the reference mark RMn is combined and imaged at the same time. The video signal Vsg corresponding to each image of the alignment marks MKn or the reference patterns FMa to FMh and the reference mark RMn imaged by the image sensor DIS is sent to the image analysis unit 206 shown in FIG.

In the present embodiment, the ejection end 404a of the optical fiber bundle 404 is set to be located at a position corresponding to the pupil surface Epp of the microscope optical system by the objective lens system OBL and the lens system Gb, and has a substantially circular outer shape. The emission end 404a formed becomes a secondary light source image in the pupil surface Epp, and telecentric epi-illumination (Koehler illumination) is performed. However, in the present embodiment, since the illumination field diaphragms FA1 to FA3 as shown in FIG. 9 above cannot be used, a reference is used to select whether or not to simultaneously image the reference mark RMn of the reference bar member RB. A liquid crystal shutter 410 on a flat surface is arranged immediately before the reference surface RBa of the bar member RB. The liquid crystal shutter 410 is individually provided for each of the detection regions ARn (that is, each of the alignment system ALGn) set on the reference bar member RB, and is driven by the measurement control unit 210 shown in FIG. The light transmittance is changed in response to the signals CCs. Each of the liquid crystal shutters 410 is fixed to a plate-shaped support frame extending in parallel with the reference bar member RB so as to be at a constant distance from the reference surface RBa of the reference bar member RB.

The transmittance of the liquid crystal shutter 410 becomes almost 0% when the voltage of the drive signal CCs is 0V, the transmittance increases as the voltage increases, and the transmittance becomes 95% or more at the nominal maximum voltage. Has various characteristics. Further, the surface of the liquid crystal shutter 410 is coated with an antireflection film so as to have a low reflectance. Therefore, at the stage of calibrating the alignment system ALGn based on the video signal Vsg from the image sensor DIS (step 302 in FIG. 12), the transmittance of the liquid crystal shutter 410 is maximized and on the reference bar member RB. The image of the reference mark RMn of the above can be detected. Further, when detecting only the alignment mark MKn on the substrate P or only the reference patterns FMa to FMh on the rotating drum DR, the transmittance of the liquid crystal shutter 410 is set to the minimum (0%).

Further, when the image of the reference mark RNn on the reference bar member RB and the image of the alignment mark MKn of the substrate P are simultaneously imaged by the image pickup device DIS, the contrast of the image of the reference mark RNn in the image pickup region DIS'is compared with the contrast. , The contrast of the image of the alignment mark MKn may be significantly reduced. In such a case, the contrast difference can be improved by adjusting the transmittance of the liquid crystal shutter 410.

As described above, in the present embodiment, the plate-type beam splitter BS1 is tilted by an angle θe with respect to the plane perpendicular to the optical axis AXs of the objective lens system OBL, and the illumination light ILb emitted from the objective lens system OBL is aligned. It was configured to face the reference bar member RB arranged in the space below the ALGn. However, when the reference bar member RB extends in the space above the alignment system ALGn, the inclination of the beam splitter BS1 with respect to the plane perpendicular to the optical axis AXs may be set in the opposite direction (−θe). In addition, the thickness of the plate-type beam splitter BS1 should be as thin as possible within a range that reduces the occurrence of various optical aberrations (ass, etc.) and has rigidity that does not cause deformation or distortion that deteriorates surface accuracy. good.

By making the surface of the plate-type beam splitter BS1 in a solid state without forming an antireflection film (AR coat), the light splitting surface (photosynthetic surface) Bsp has an appropriate reflectance depending on the angle θe. Can be made. Further, the angle θe is determined by the angle 2θe (arrangement of the reference bar member RB in the XZ plane) formed by the optical axis AXs and the optical axis AXs', but when the angle θe becomes, for example, 45 ° or more, the reference bar member Since the intensity of the illumination light ILb toward the RB increases and the intensity of the illumination light ILb toward the substrate P decreases extremely, the angle θe is in the range of 0 ° <θe <45 °, more preferably 5 °. It is preferable to set the range to ≤θe≤30 °. Further, the thickness of the plate-type beam splitter BS1 may be 1 mm or less, for example 0.1 mm, and a structure capable of adjusting the angle θe may be provided.

[Modification 3]
FIG. 25 is an optical layout diagram showing a modification of the beam splitter BS1 (synthetic optical member) in the third embodiment shown in FIG. 24, and the optical axis AXs, the plane mirror Mb, and the like of the objective lens system OBL are shown. It is provided in the same manner as the arrangement in the XYZ coordinate system of FIG. 24. In this modification, as shown in FIG. 25, the beam splitter BS1 is formed in a cube shape, and the prism block PSMa made of quartz located on the objective lens system OBL side and the prism block PSMb made of quartz located on the plane mirror Mb side are attached. By combining, the overall shape of the cross section along the plane parallel to the XZ plane is formed to be trapezoidal (pentagonal). The bonding surface Bsp of the prism block PSMa and the prism block PSMb functions as an optical dividing surface, and is configured to be tilted in the XZ plane by an angle θe with respect to the plane perpendicular to the optical axis AXs. The illumination light ILb emitted from the objective lens system OBL is incident from the surface BS1a of the prism block PSMa, and the illumination light ILb transmitted through the bonded surface (divided surface) Bsp is emitted from the surface BS1b of the prism block PSMb to be a flat mirror. Reach Mb.

The surface BS1a of the prism block PSMa and the surface BS1b of the prism block PSMb are parallel to each other and set perpendicular to the optical axis AXs. Further, the illumination light ILb reflected by the dividing surface Bsp is emitted from the surface BS1c of the prism block PSMA along the optical axis AXs' that is set perpendicular to the reference surface RBa of the reference bar member RB. The surface BS1c of the prism block PSMa is formed so as to be perpendicular to the optical axis AXs'. The angle formed by the optical axis AXs and the optical axis AXs'is 2θe because it is a double angle of the angle θe of the dividing surface Bsp. 2θe). Also in this modification, a dielectric film is formed on the bonded surface (divided surface) Bsp so that the reflectance of the illumination light ILb on the bonded surface (divided surface) Bsp is about 10 to 30%. ..

According to the above modified example, the alignment mark MKn of the substrate P illuminated by the illumination light ILb or the reflected light (imaging light beam Bma) from the reference patterns FMa to FMh of the rotating drum DR is the light of the beam splitter BS1. Since it is incident on the objective lens system OBL through the planes BS1b and BS1a perpendicular to the axes AXs, it is possible to reduce the occurrence of various optical aberrations at the time of imaging of the alignment mark MKn and the reference patterns FMa to FMh. Similarly, the reflected light (imaging light beam Bmr) from the reference mark RMn of the reference bar member RB illuminated by the illumination light ILb is incident from the plane BS1c perpendicular to the optical axis AXs'of the beam splitter BS1 and is incident on the optical axis AXs. Since it is incident on the objective lens system OBL through the plane BS1a perpendicular to the image, the occurrence of various optical aberrations at the time of imaging of the reference mark RMn can be reduced.

[Fourth Embodiment]
FIG. 26 is a diagram showing the configuration of the reference bar member RB according to the fourth embodiment, FIG. 26A shows the configuration of the reference bar member RB on the reference surface RBa, and FIG. 26B is the reference in FIG. 26A. The CC-CC arrow cross section of the bar member RB is shown, and FIG. 26C shows an example of the configuration of the reference mark RMn formed on the reference surface RBa. In the present embodiment, the reference mark RMn on the reference bar member RB is self-luminous with adjustable illuminance. Therefore, as shown in FIG. 26B, an optical fiber bundle 450 for guiding the illumination light ILh for self-luminous light is connected inside a metal or ceramic square lumber serving as a base material RBo of the reference bar member RB. NS.

As shown in FIG. 20, depending on the configuration in which the seven alignment systems ALG1 to ALG7 are provided, also in FIG. 26A, each of the positions in the Y direction corresponding to the detection regions AD1 to AD7 of the alignment systems ALG1 to ALG7. Reference marks RM1 to RM7 (center point CRn) are arranged therewith. A thin quartz plate RBg is provided extending in the Y direction on the surface of the base material RBo of the reference bar member RB on the beam splitter BS1 side, and a low reflectance chromium layer (light-shielding layer) is provided on the surface of the quartz plate RBg. ) Is deposited to form a reference surface RBa. As shown in FIG. 26C, at each position of the reference mark RMn on the reference surface RBa (light-shielding layer), a rectangular transmission window WDn (n = 1 to 7) having dimensions corresponding to each detection region ADn of the alignment system ALGn. ) Is formed, and the reference mark RMn by the light-shielding layer is arranged inside the).

The base material RBo on the back surface side of the quartz plate RBg is formed with an opening RBz having a columnar hole with a diameter so as to include each of the transmission windows WDn, and an optical fiber bundle 450 is injected into the opening RBz. A lens member 452 is provided which injects the illumination light ILe projected from the end and uniformly illuminates the entire individual transmission window WDn from the back surface side (quartz plate RBg side). Further, the injection end side of the optical fiber bundle 450 and the lens member 452 are fixed in the opening RBz via a heat insulating resin member 454 formed in a circular tubular shape. The lens member 452 is formed at the ejection end of the optical fiber bundle 450, and the transmission window WDn is illuminated from the back side by Koehler illumination using a circular set of a large number of point light sources as a secondary light source. The optical axis AXi of the lens member 452 shown in FIG. 26B is coaxial with the optical axis AXs of the objective lens system OBL described in each of FIGS. 5, 23, and 24 above via the beam splitter BS1 within a tolerance. Is set to be.

In the present embodiment, the illumination light ILh from the optical fiber bundle 450 is irradiated into the transmission window WDn via the lens member 452 provided in the base material RBo of the reference bar member RB. The enlarged image RNn'of the reference mark RNn observed by DIS appears as a black pattern on a bright background (white). Further, the illumination light ILh can be supplied from the light source unit in the illumination system ILU (see FIG. 6) that supplies the illumination light ILb for the alignment system ALGn, but another illumination system ILU'is provided to provide the illumination light. The illuminance of ILh may be made variable. When another illumination system ILU'is used, the illuminance of the illumination light ILh is adjusted regardless of the illuminance of the illumination light ILb for the alignment system ALGn, or only when the alignment system ALGn is calibrated (step 302 in FIG. 12). Is irradiated with the illumination light ILh, and the illumination light ILh can be turned off during the period in which the alignment system ALGn sequentially detects the alignment mark MKn on the substrate P.

Further, the wavelength characteristics of the illumination light ILb for the alignment system ALGn and the illumination light ILh for the reference mark RMn (transmission window WDn) of the reference bar member RB may be different. For example, the illumination light ILb for the alignment system ALGn is non-photosensitive to the photosensitive layer on the substrate P and has a broad wavelength band (for example, 450 nm or more) so as to be suitable for detecting the alignment mark MKn on the substrate P. The illumination light ILh for the reference mark RMn (transmission window WDn) of the reference bar member RB is the light having 700 nm), and the image is obtained by reducing the chromatic aberration generated in the imaging optical system such as the objective lens system OBL of the alignment system ALGn. It is possible to make monochromatic light (non-photosensitive) that can enhance the contrast.

Further, in the present embodiment, as shown in FIG. 26C, a transparent transparent window WDn is provided in the reference surface RBa formed by the light-shielding layer, and a reference mark RMn by the light-shielding layer is formed in the transparent window WDn. However, the transparent reference mark RMn may be formed in the reference surface RBa by the light-shielding layer without providing the transparent window WDn. In this case, the enlarged image RNn'of the reference mark RMn observed by the image sensor DIS of the alignment system ALGn appears as a bright pattern (white) on a dark background (black). Therefore, even while the alignment system ALGn detects the alignment mark MKn on the substrate P, the enlarged image RNn'of the reference mark RNn is high in the entire image of the surface of the substrate P appearing in the imaging region DIS'. It can be displayed in contrast.

Further, the illuminance of the illumination light ILb for the alignment system ALGn can be adjusted separately from the illuminance of the illumination light ILh for the reference mark RMn of the reference bar member RB, so that the reflection of the surface (alignment mark MKn) of the substrate P can be adjusted. When the rate is low, the illuminance of the illumination light ILb for the alignment system ALGn is increased, and when the reflectance of the surface (alignment mark MKn) of the substrate P is high, the illuminance of the illumination light ILb for the alignment system ALGn is decreased independently. It will be possible. Therefore, even if the reflectance (or absorption rate) of the surface of the substrate P fluctuates, the deterioration of the measurement accuracy of the alignment mark MKn on the substrate P during image analysis is suppressed, and stable position measurement is possible. Become. The reference bar member RB of FIG. 26 according to the present embodiment can be used as it is in combination with any of the alignment system ALGn shown in each of FIGS. 23, 24, and 25 above.

[Fifth Embodiment]
FIG. 27 shows a schematic configuration of a pattern drawing device using a digital mirror device (DMD) as a maskless exposure device. The Z axis of the Cartesian coordinate system XYZ is set in the direction of gravity and is perpendicular to the Z axis. The XY plane is set to a horizontal plane. In FIG. 27, the substrate P as an object to be exposed is mounted on a moving stage (not shown) that translates in one dimension (X direction) or two dimensions (X direction and Y direction) along a plane parallel to the XY plane. Placed. In the present embodiment, the substrate P is a single-wafer flat glass substrate, a plastic substrate, or a metal substrate, but a resin sheet such as PET (polyethylene terephthalate) film, PEN (polyethylene naphthalate) film, or polyimide film is used. There may be.

Also in this embodiment, the odd-numbered drawing units (pattern forming mechanism) U1, U3, U5 ... Arranged in a row at regular intervals in the Y direction and the odd-numbered drawing units are predetermined in the X direction. The even-numbered drawing units (pattern forming mechanism) U2, U4, U6, etc., which are arranged at regular intervals in the Y direction, are provided. In FIG. 27, on the substrate P, the projection areas IA1, IA3, IA5 ... By the odd-numbered drawing units U1, U3, U5 ..., And the even-numbered drawing units U2, U4, U6 ... The projection areas IA2, IA4, IA6, ... The odd-numbered drawing units U1, U3, U5 ... And the even-numbered drawing units U2, U4, U6 ... Are arranged symmetrically with respect to the central plane CPo parallel to the YZ plane when viewed in the XZ plane. At the same time, the projection area IAn is shifted in the Y direction by about the dimension of the Y direction. Since the configurations of the odd-numbered and even-numbered drawing units Un (n = 1, 2, 3, ...) Are the same, the configuration of the drawing unit U1 will be described in detail here as a representative.

The drawing unit U1 includes a support frame 550 made of metal having a low thermal expansion coefficient that supports the entire main member, an illumination optical system (lens barrel) 551 that injects an illumination beam LB1 (monochromatic light in the ultraviolet wavelength range) for exposure. A reflection mirror 552 that reflects the illumination beam LB1 having a uniform illumination distribution by the illumination optical system 551, a DMD unit 553 that is irradiated by the illumination beam LB1 from the reflection mirror 552, and a large number of micro mirrors of the DMD unit 553. A projection optical system (lens barrel) that projects a reduced image of a dynamic pattern into the projection area IA1 on the substrate P by injecting a drawing beam that is modulated by sequentially changing each angle according to the pattern data. It is composed of 554 and.

The illumination beam LB1 is incident on the illumination optical system (lens cylinder) 551 from a laser light source via an optical fiber bundle, and is made into a uniform illuminance distribution by a fly-eye lens, a condenser lens, etc. in the illumination optical system 551, and the DMD unit 553. Illuminate the Koehler. When viewed in the XY plane, the reflective surface of the DMD unit 553 is tilted by a predetermined angle for joint exposure with the projection region IA2 by the drawing units U2 adjacent to each other in the Y direction. The projection optical system (lens barrel) 554 is a bilateral telecentric connection composed of a plurality of lens elements in a straight cylinder shape so that the entire reflection surface of the DMD unit 535 is reduced and imaged in the projection region IA1 on the substrate P. It is configured as an image optical system.

On the other hand, a plurality of odd-numbered alignment systems ALG1, ALG3 ... Are arranged on the -X direction side of the odd-numbered drawing units U1, U3, U5 ..., And the even-numbered drawing units U2, U4, U6 ... On the + X direction side of ..., a plurality of even-numbered alignment systems ALG2, ALG4 ... Are arranged. The odd-numbered alignment systems ALG1, ALG3 ... Are provided so that the respective detection regions AD1, AD3 ... Are located in a row at predetermined intervals in the Y direction, and the even-numbered alignment systems ALG2, ALG4 ... ... Are provided so that the respective detection areas AD2, AD4 ... Are located in a row at predetermined intervals in the Y direction. The odd-numbered alignment systems ALG1, ALG3 ... And the even-numbered alignment systems ALG2, ALG4 ... Are arranged symmetrically with respect to the central plane CPo when viewed in the XZ plane.

Since the configurations of the alignment system ALGn (n = 1, 2, 3, 4, ...) Are the same, the configuration of the alignment system ALG1 will be described in detail here as a representative. The alignment system ALG1 is composed of an illumination system ILU, an image pickup element DIS, an objective lens system OBL, and a rectangular parallelepiped beam splitter BS1 as in FIGS. 6 and 9, and the optical axis AXs of the objective lens system OBL is a substrate. It is set perpendicular to the surface of P (parallel to the Z axis). On the side of the beam splitter BS1 arranged in the optical path between the objective lens system OBL and the substrate P, the odd-numbered drawing units U1, U3, U5 ... Are fixed to the support frames 550 via mounting brackets. The first reference bar member RB to be split is arranged. The first reference bar member RB is elongated so as to penetrate in the Y direction near the odd-numbered alignment systems ALG1, ALG3, ALG5 ..., And the reference mark RMn of the first reference bar member RB is formed. The reference plane RBa to be made is set parallel to the YZ plane. Therefore, the optical axis AXs'between the beam splitter BS1 and the reference bar member RB is set to be parallel to the XY plane and parallel to the X axis.

Similarly, in the even-numbered alignment systems ALG2, ALG4, ALG6 ..., The beam splitter BS1 is arranged in the optical path between the objective lens system OBL and the substrate P, and the even-numbered drawing unit U2 is located on the side thereof. , U4, U6 ... A second reference bar member RB fixed to the support frame 550 via a mounting bracket is arranged.

FIG. 28 shows the projection region IAn (n = 1, 2, 3 ...) by each of the drawing units Un (n = 1, 2, 3 ...) And the alignment system ALGn (n = 1, 2, 3 ...). The arrangement of the detection regions ADn (n = 1, 2, 3 ...) By each of (...) in the XY plane is shown. By arranging the DMD unit 553 at an angle in the XY plane, each of the rectangular projection regions IAn is arranged so as to be inclined by an angle θdm with respect to a line parallel to the Y axis. Further, each of the ends of the projection region IAn on the Y direction side has overlapping regions (joint portions) OL12, OL23, OL34, OL45, OL56 ... Arranged to be formed. Further, the odd-numbered reference mark RM1 is placed on the first reference bar member RB corresponding to the arrangement of the detection regions AD1, AD3 ... Of the odd-numbered alignment systems ALG1, ALG3 ... In the Y direction. , RM3 ... Are formed, and corresponding to the arrangement of the detection regions AD2, AD4 ... Of the even-numbered alignment systems ALG2, ALG4 ... In the Y direction, on the second reference bar member RB. The even-numbered reference marks RM2, RM4, ... Are formed. In the present embodiment, each of the rectangular projection regions IAn (n = 1, 2, 3 ...) With the Y direction as the long side, or a plurality of projection regions IAn (n = 1, 2, 3, ...). A pattern-forming region is formed by a rectangular region including the entire).

With the above configuration, also in the present embodiment, as in each of the previous embodiments, the relative positional relationship due to the drift of the plurality of alignment systems ALGn is exposed during the exposure operation with reference to the reference bar member RB. Even so, it is possible to monitor sequentially. In the present embodiment, when the pattern is exposed on the substrate P, the substrate P is moved by the moving stage at a constant speed, for example, in the + X direction. In that case, the alignment mark MKn on the substrate P is mainly detected by the odd-numbered alignment systems ALG1, ALG3 ... (Detection regions AD1, AD3 ...), And the odd-numbered and even-numbered based on the detection result. The position of the pattern image projected in each projection area IAn of the drawing unit Un and the projection timing (the timing of sending out the pattern data applied to the DMD unit 553) are dynamically fine-tuned. Further, when pattern exposure is performed on the substrate P, the moving stage may be moved in the −X direction at a constant speed. In that case, the alignment mark MKn on the substrate P is mainly detected by the even-numbered alignment systems ALG2, ALG4 ... (Detection areas AD2, AD4 ...) The position and projection timing of the pattern image projected in each projection area IAn of the even-numbered drawing unit Un are dynamically fine-tuned.

[Modification example 4]
The DMD unit 553 shown in FIGS. 27 and 28 finely moves each of a large number of micro mirrors in a direction perpendicular to the reflecting surface to make a difference in the height of the reflecting surface between the adjacent micro mirrors. It may be a spatial light modulation element (SLM) that gives (give a phase difference). Further, the pattern drawing apparatus having a plurality of alignment systems is not limited to the maskless exposure apparatus, and may be an inkjet printer apparatus. In the inkjet method, a nozzle unit (also called a drawing unit) as a pattern forming mechanism having an ejection surface in which a large number of fine pores for ejecting fine droplets of ink are regularly arranged is formed by a substrate P and an ejection surface. Arranged at regular intervals. When drawing (printing) a pattern, while moving the substrate P in one direction (secondary scanning direction), liquid is selectively liquid on the substrate P from each of a large number of minute holes in the nozzle unit (drawing unit) based on pattern data. Drops are ejected.

Even in an inkjet printer device, it is necessary to draw (overprint) a new pattern in a precisely positioned state with respect to the pattern layer already formed on the substrate P, and when the size of the substrate P becomes large. In order to accurately measure the deformation (expansion and contraction in the plane and distortion), a plurality of alignment systems for detecting each position of the alignment marks MKn formed at a plurality of positions on the substrate P are required. Therefore, by providing the reference bar member RB as described in each of the above embodiments and modifications and the alignment system ALGn having the beam splitter BS1, pattern drawing (overprint printing) is continuously performed over a long period of time. ), It is possible to suppress the occurrence of a superposition error that may occur due to the drift of the alignment system ALGn. In this modification, the area of the droplet ejection surface by one nozzle unit (drawing unit), or when having a plurality of nozzle units, the entire droplet ejection surface of each of the plurality of nozzle units is covered. A pattern-forming region is formed by the region including the region.

[Modification 5]
FIG. 29 is a diagram illustrating a modified example relating to the configuration of the alignment system ALGn, and FIG. 29A shows a modified example based on the alignment system ALGn shown in FIG. 24 above. In this modification, the illumination light ILb for the alignment system ALGn has a wide band wavelength distribution, and the beam splitter BS1 (synthetic optical member) arranged between the objective lens system OBL and the substrate P is placed on the surface of a dielectric multilayer film. Is formed to form a dichroic mirror having wavelength selectivity. FIG. 29B is a graph showing an example of the wavelength selection characteristics of the beam splitter BS1, where the horizontal axis represents the wavelength λ (nm) and the vertical axis represents the magnitudes of transmittance and reflectance. In this modification, aberration correction is performed so that the imaging optical system from the objective lens system OBL of the alignment system ALGn to the image sensor DIS can obtain good imaging characteristics within the wavelength range of the chromatic aberration correction range shown in FIG. 29B. It is assumed that it has been done.

Further, the illumination system ILU that supplies the illumination light ILb for the alignment system ALGn includes a light source HLS such as a halogen lamp having light emission intensity over a wavelength band of about 400 nm to 700 nm, and a wavelength distribution and wavelength of light from the light source HLS. In order to make the intensity adjustable, a wavelength selection unit WLC in which a plurality of wavelength filters are interchangeably housed and a light source system Gk are provided, and the light source system Gk is provided as in FIG. 24. Illumination light ILb is applied to the incident end 404b of the optical fiber bundle 404. The emission end 404a of the optical fiber bundle 404 is set at the pupil position of the objective lens system OBL of the alignment system ALGn, and the illumination light ILb is the alignment mark MKn on the substrate P or the reference via the beam splitter BS1 (dichroic mirror). The reference mark RMn of the bar member RB is illuminated.

As an example, the beam splitter BS1 (dichroic mirror) of FIG. 29A has a transmittance characteristic and a reflectance characteristic crossover at around 500 nm in a wavelength region longer than 420 nm, which is a resist photosensitive region, as shown in FIG. 29B. It has wavelength selection characteristics such as. That is, the reflectance is 5% or more for light having a wavelength component shorter than about 550 nm, and the transmittance is 5% or more for light having a wavelength component longer than about 480 nm. It has such characteristics. Therefore, of the illumination light ILb emitted from the objective lens system OBL and applied to the beam splitter BS1 (dichroic mirror), the light having a wavelength component shorter than about 550 nm is efficiently reflected and the reference bar member RB. Illuminates the reference mark RMn. Further, among the illumination light ILb, light having a wavelength component having a wavelength longer than about 480 nm is efficiently transmitted and illuminates the alignment mark MKn on the substrate P. The wavelength selection unit WLC typically selects the wavelength so that the illumination light ILb has an intensity distribution over the wavelength band of the chromatic aberration correction region.

As described above, in this modification, the illumination light ILb that illuminates the reference mark RMn of the reference bar member RB is set to the light of the component on the short wavelength side in the chromatic aberration correction range, and illuminates the alignment mark MKn on the substrate P. The wavelength of the illumination light ILb is selected so as to be set to the light of the component on the long wavelength side within the chromatic aberration correction range. Therefore, even when it becomes necessary to continuously irradiate the reference bar member RB with the illumination light ILb, the light component on the long wavelength side, which tends to raise the temperature of the reference bar member RB, is reduced, so that the reference bar member RB has a reduced light component. Thermal deformation of the member RB itself can be suppressed.

[Modification 6]
Further, according to the above modification 5, since the beam splitter BS1 (dichroic mirror) has a wavelength selection characteristic, the illuminance of the illumination light ILb that illuminates the alignment mark MKn on the substrate P and the reference bar member RB The illuminance of the illumination light ILb that illuminates the reference mark RMn of the above can be adjusted individually (independently). Specifically, as shown in FIG. 30, two solid light sources (LEDs and the like) LD1 and LD2 having different emission wavelength ranges, lens systems GS1 and GS2, a mirror Mc, a beam splitter BS3 for beam synthesis, and the like. Illumination system ILU including a condensing lens system Gk that condenses the illumination light ILb at the incident end of the optical fiber bundle 404 shown in FIG. 29, and a control unit LCU that can individually adjust the emission intensity of the solid-state light sources LD1 and LD2. Is provided.

The beam ILb1 from the solid-state light source LD1 converted into a parallel light flux by the lens system GS1 includes, for example, an emission spectrum having a single or multiple center wavelengths between wavelengths of 420 nm and 500 nm, and is converted into a parallel light flux by the lens system GS2. The beam ILb2 from the solid-state light source LD2 to be converted is set to include, for example, an emission spectrum having a plurality of center wavelengths between wavelengths of 500 nm and 630 nm. The beam ILb1 from the lens system GS1 reflected by the mirror Mc and the beam ILb2 from the lens system GS2 form an illumination light ILb coaxially combined by the beam splitter BS3 and enter the condenser lens system Gk to form an optical fiber bundle. It is incident on the beam splitter BS2 in FIG. 29 via 404.

According to the illumination system ILU as shown in FIG. 30, since the intensity of the beams ILb1 and ILb2 from each of the solid light sources LD1 and LD2 can be individually adjusted by the control unit LCU, the alignment mark MKn on the substrate P can be adjusted by the image sensor DIS. When the magnified image of the above and the magnified image of the reference mark RMn of the reference bar member RB are simultaneously imaged, the contrast and brightness of each magnified image can be adjusted to an appropriate balance in the captured image. In particular, when the reflectance of the surface of the substrate P with respect to the illumination light ILb (ILb2) is generally low and the magnified image of the alignment mark Mkn is dark, the emission intensity of the solid light source LD2 (beam ILb2) is increased and the solid light source LD1 (beam ILb1) is increased. By lowering the light emission intensity of the light source and increasing the gain (amplification rate) of the video signal Vsg from the image pickup element DIS, the lightness and darkness of the observed image can be balanced. Further, when the image sensor DIS is capable of capturing a color image, the alignment marks MKn (or the reference patterns FMa to FMh on the rotating drum DR) on the substrate P and the reference marks RMn on the reference bar member RB are formed. Since it can be identified and detected by color, it is possible to reduce erroneous detection when the image of the alignment mark MKn and the image of the reference mark RMn appear close to each other in the imaging region DIS'.

In order to set each of the beams ILb1 and ILb2 to a desired wavelength band, the solid-state light sources LD1 and LD2 are set to high-intensity white light emitting LEDs, and in each optical path of the beams ILb1 and ILb2 in front of the beam splitter BS3 for synthesis. A wavelength selection filter (bandpass filter) may be provided in the. Further, in order to reduce the light loss in the beam splitter BS3 for synthesis, the beam splitter BS3 is used as a polarizing beam splitter, and the beam ILb1 is linearly polarized so as to transmit 90% or more of the polarization separation surface of the beam splitter BS3, and the beam ILb2 A wavelength plate for linearly polarized light that is reflected by 90% or more on the polarization splitting surface of the beam splitter BS3 may be provided after each of the lens systems GS1 and GS2.

In this way, when the illumination light ILb can be imparted with polarization characteristics, the beam splitter BS1 (synthetic optical member) of the alignment system ALGn shown in FIGS. 9, 23, 24, and 25 above also has an optical splitting surface (composite optical member). Alternatively, it can be a polarization beam splitter in which a dielectric multilayer film having polarization selectivity is formed on Bsp (photosynthesis surface). Further, the illumination light ILb that is reflected by the beam splitter BS1 and illuminates the reference mark RMn of the reference bar member RB is light in the near infrared wavelength range from an LED or the like, and the alignment mark MKn on the substrate P is set from the beam splitter BS1. The illumination light ILb to be illuminated may be light in a wide wavelength range (so-called randomly polarized white light) avoiding the photosensitive wavelength range (for example, an ultraviolet wavelength range of 360 nm or less) of the light sensitive layer on the substrate P. ..

[Modification 7]
FIG. 31 shows a modification relating to the configuration of the beam splitter BS1 of the alignment system ALGn arranged between the objective lens system OBL and the substrate P (outer peripheral surface DRs of the rotating drum DR) and the arrangement direction of the reference bar member RB. It is a figure which shows. The beam splitter BS1 (synthetic optical member) of this modification has a configuration in which two prism blocks PSMa and PSMb are bonded together and an optical splitting surface Bsp is formed at the contact interface thereof, as in the configuration of FIG. 25 above. .. However, a part of the illumination light ILb that is incident on the prism block PSMA from the objective lens system OBL along the optical axis AXs and is reflected by the light dividing surface Bsp is a total reflection surface (mirror surface) of the surface BS1c of the prism block PSMa. As a result, the surface BS1c is bent in the direction of the optical axis AXs passing through the objective lens system OBL. In this modification, the optical axis AXs'extending from the optical division surface Bsp is bent at the surface BS1c (total reflection surface) and set so as to be orthogonal to the optical axis AXs.

The Cartesian coordinate system XYZ in FIG. 31 is set to be the same as in FIGS. 23 and 24, and the optical axis AXs between the objective lens system OBL and the plane mirror Mb is tilted with respect to the XY plane. The beam splitter BS1 is formed in a pentagonal shape as a whole when viewed in the XZ plane, and the plane BS1a on the prism block PSMa side facing the objective lens system OBL and the plane BS1b on the prism block PSMb side facing the plane mirror Mb are It is set parallel to each other and perpendicular to the optical axis AXs. The optical division surface Bsp between the surface BS1a and the surface BS1b, which is the junction surface of the prism blocks PSMa and PSMb, is tilted by an angle θe with respect to the surface perpendicular to the optical axis AXs, as in FIG. 25 above. Is set. Therefore, at the position of the optical division surface Bsp, the angle formed by the optical axis AXs and the optical axis AXs'is an angle 2θe in the XZ surface.

The optical axis AXs'bent at the surface BS1c (mirror surface) of the prism block PSMa and extends in the + Z direction side intersects the optical axis AXs at a right angle, and then passes through the upper surface BS1d of the prism block PSMa and passes through the reference bar. It reaches the reference surface RBa (center point of the reference mark RMn) of the member RB. When viewed in the XZ plane, the plane BS1d forms a right angle with each of the planes BS1a and BS1b, is set parallel to the optical axis AXs, and is further set parallel to the reference plane RBa of the reference bar member RB. Therefore, the inclination of the surface BS1c (mirror surface) in the XZ surface is set so that the angle formed by the optical axes AXs' folded back at the position of the surface BS1c (mirror surface) is an angle (90 ° -2θe). There is. The angle θe can be theoretically set in the range of 0 ° <θe <45 °, but the thickness and numerical aperture of the illumination light ILb, the thickness of the imaging light beam, and the numerical aperture of the objective lens system OBL, respectively. It is set in the range of 5 ° ≤ θe ≤ 35 ° due to the limitation of the arrangement relationship of the members. This also applies to the arrangement of the beam splitter BS1 in FIGS. 24 and 25. In the case of the beam splitter BS1 of FIG. 31, when the angle θe of the optical splitting surface Bsp is set to 22.5 °, the intersection of the optical axis AXs and the optical axis AXs' and the optical axis AXs in the optical splitting surface Bsp pass through. The three points, the point and the point where the optical axis AXs' in the plane BS1c (mirror plane) is folded back, form each vertex of a right-angled isosceles triangle in the XZ plane.

This modification can be applied when it is difficult to arrange the reference bar member RB in the space below the alignment system ALGn (objective lens system OBL, beam splitter BS1, etc.). Further, as in the beam splitter BS1 of FIG. 31, the beam splitter BS1 has an orthogonal relationship between the optical axis AXs passing through the objective lens system OBL and the optical axis AXs' from the surface BS1c toward the reference bar member RB. Even if the whole is tilted in the XZ plane and attached to the support bracket 400 (see FIG. 24), the optical axis AXs from the plane BS1b of the beam splitter BS1 toward the plane mirror Mb and the reference bar from the plane BS1d of the beam splitter BS1. The optical axes AXs' toward the member RB are only slightly parallel-shifted in the lateral direction in the XZ plane, and the optical axes AXs and AXs' are not tilted. That is, the configuration is such that the occurrence of telesen error is suppressed. Further, in this modification as well, a dielectric multilayer film for wavelength selection or polarization separation may be formed on the optical division surface Bsp which is the junction surface between the prism block PSMa and the prism block PSMb.

In each of the embodiments and modifications described above, the reference bar member RB uses a base material (RBo) as a material having a low coefficient of thermal expansion (Invar alloy made of Fe-36Ni, Kovar alloy made of Fe29Ni-17Co, HfW). _{By using 2} O ₈ (or ZrW ₂ O ₈ ) and Mg WO ₄ mixed sintering material, quartz, kovarite ceramics, glass ceramics, etc.), the dimensional change in the longitudinal direction due to the temperature change of the environment and heat conduction Can be ignored. However, the reference bar member RB extends in a rod shape along the arrangement direction (Y direction) of the plurality of alignment systems ALGn. Therefore, depending on the relationship between the dimension of the reference bar member RB in the longitudinal direction (Y direction) and the dimension of the cross-sectional shape, it can be ignored as compared with the position detection accuracy of the alignment mark MKn by the alignment system ALGn due to its own weight. It may be deformed to the extent that it does not (cause bending or bending). The deformation state and the amount of deflection of the reference bar member RB due to its own weight can be uniquely identified by the strength of materials calculation based on the support structure in the longitudinal direction of the reference bar member RB and physical conditions. ..

FIG. 32 shows the difference in the deformed state depending on the support structure of the reference bar member RB, and FIG. 32A shows the structure in which the vicinity of both ends of the reference bar member RB in the longitudinal direction is supported from below by the beam FJ1 which makes line contact with each other. 32C shows a structure in which the vicinity of one end of the reference bar member RB in the longitudinal direction is supported from below by a beam FJ1 that makes line contact, and the other end is fastened (fixed) to the device frame FJ2. FIG. 32C shows the reference. A structure is shown in which both ends of the bar member RB in the longitudinal direction are fastened (fixed) to the device frame FJ2. As shown in FIGS. 32A to 32C, even if the base material and the dimensions of the reference bar member RB are the same, the bending states SV1, SV2, and SV3 change depending on the respective support structures.

The amount of deformation due to the bending states SV1, SV2, and SV3 can be easily calculated by strength of materials calculation, and the error in the relative positional relationship of the reference mark RMn formed on the reference surface RBa of the reference bar member RB is precise in advance. Desired. The error in the relative positional relationship is the center point CCn (center point of the imaging region DIS') of the detection region ADn by each of the plurality of alignment systems ALGn at the time of the alignment system calibration in step 302 described in FIG. It is introduced as a correction value when determining the relative positional relationship of. As a result, each of the plurality of alignment systems ALGn is calibrated with high accuracy even if the reference bar member RB is bent. As for the support structure of the reference bar member RB, in addition to the configuration shown in FIG. 32, only the vicinity of the center of the reference bar member RB in the longitudinal direction is fastened to a part of the device frame FJ2 by screwing, and both ends are in the longitudinal direction. It may be a structure that supports its own weight so as not to be restrained by. Even in that case, the bending state of the reference bar member RB can be specified by the strength of materials calculation.

[Modification 8]
In each of the above embodiments and modifications, the cross-sectional shape parallel to the XZ plane of the reference bar member RB is a rectangle (rectangle), but the cross-sectional shape is a polygon such as a triangle or a pentagon, or a circle (surrounding). A part can be cut into a flat surface) or an L-shaped angle can be formed. Further, in order to reduce the weight of the reference bar member RB, a hollow structure or a lightening structure may be used as long as the required rigidity can be obtained. Further, as described with reference to FIG. 8, when the reference bar member RB is held by the

support members

103A and 103B (103B is not shown) on both ends in the Y direction, the reference bar member RB is an L-shaped angle member. It may be attached to.

FIG. 33 shows a modified example of the support structure of the reference bar member RB by the

support members

103A and 103B described in FIG. 8 in the range from the vicinity of the center of the reference bar member RB to the support member 103B on the + Y direction side. It is a perspective view, and the Cartesian coordinate system XYZ is set in the same manner as in FIG. Further, the configuration of the reference bar member RB itself is the same as that of FIG. 10 or 26 above, and here, seven reference marks RM1 to RM7 corresponding to each of the seven alignment systems ALG1 to ALG7 are on the reference surface RBa. Is formed in. The

support members

103A and 103B are formed of a metal material having a low coefficient of thermal expansion, and the angle members 108 having an L-shaped cross section are bridged to the

support members

103A and 103B so as to be parallel to the Y axis. It is fixed. The reference bar member RB having a rectangular cross-sectional shape is locked on the angle member 108 by a fixing portion 109 provided near the center. Both ends of the reference bar member RB in the Y direction are locked so as not to generate a binding force in the Y direction.

By attaching the reference bar member RB to the device frame (

support members

103A, 103B) via the angle member 108 in this way, the coefficient of thermal expansion is extremely small as the base material of the reference bar member RB, but the amount of deflection due to its own weight. It is possible to use a material that has a large coefficient or has a low rigidity and is easy to shear. With the support structure as shown in FIG. 33, even with such a material, even if the cross-sectional area in the XZ plane is small, the reference bar having a dimension of several tens of centimeters or more (for example, 30 cm or more) in the Y direction. It can be a member RB. The angle member 108 may have a cross-sectional shape other than the L-shape on a plane parallel to the XZ plane, and may be a simple rectangle (rectangle, trapezoid, parallelogram, etc.), a triangle, or a semicircle. Further, a precise temperature sensor (for example, a measurement resolution of 0.2 ° C. or less) may be provided at each of a plurality of locations in the longitudinal direction of the angle member 108 to monitor changes in the temperature distribution of the reference bar member RB.

[Other variants]
In each of the above embodiments and modifications, the four alignment systems ALG1 to ALG4 or the seven alignment systems ALG1 to 7 over the range of the short length LPy (see FIGS. 7 and 20) of the substrate P in the Y direction. Although ALG7 is provided, when the expansion and contraction of the substrate P itself and the distortion deformation in the plane are extremely small, or when the short length LPy is small, the number of drawing units Un for maskless exposure and inkjet printing is also 1 to 1. When only about 2 is required, the alignment marks MKn on the substrate P may be formed only on both ends of the substrate P in the Y direction. In such a case, since the alignment system ALGn is also arranged at two locations separated in the Y direction, the reference mark (reference index mark) RMn on the reference bar member RB is also the detection region ADn (ARn) of the two alignment system ALGn. It may be formed at two positions in the Y direction corresponding to each of the above.

On the reference surface RBa of the reference bar member RB, the rectangular detection region ARn set to include the reference mark (reference index mark) RMn is aligned except in the case of the self-luminous method described with reference to FIG. 26 above. It is illuminated by a part of the illumination light ILb from the objective lens system OBL of the system ALGn, and the reflected light from the detection region ARn is captured by the image sensor DIS via the objective lens system OBL. Therefore, it is necessary to give a large difference between the reflectance of the reference mark RMn itself in the detection region ARn and the reflectance of the background portion (reference surface RBa itself) around the reference mark RNn to enhance the contrast of the image at the time of imaging. desirable. When the reflectance of the reference surface RBa itself is high (for example, when it is 40% or more) due to the base material of the reference bar member RB, the reflectance of the reference mark RMn itself becomes a sufficiently low value (for example, 5% or less). On the contrary, when the reflectance of the reference surface RBa itself is low (for example, when it is 20% or less), the reflectance of the reference mark RMn itself becomes a sufficiently high value (for example, 60% or less). Is formed to be.

Further, the illumination light ILb projected from each objective lens system OBL of the alignment system ALGn is randomly polarized, and the cube-type or plate-type beam splitter BS1 is unpolarized, thereby irradiating the substrate P with illumination light. ILb can also be randomly polarized. Further, the cube-type or plate-type beam splitter BS1 is made into a polarized light type, and the illumination light ILb of linearly polarized light (P-polarized light) toward the substrate P and the illumination light ILb of linearly polarized light (S-polarized light) toward the reference bar member RB by polarization division are used. A rotatable wavelength plate so that the intensity ratio of the P-polarized light component and the S-polarized light component of the illumination light ILb incident on the objective lens system OBL from the illumination system ILU or the ellipticity of circularly polarized light can be changed. Etc. may be provided to adjust the intensity of the illumination light.

In the alignment system ALGn described in each embodiment and each modification, an image detection method is used to obtain an image of the alignment mark MKn on the substrate P and the reference mark RMn on the reference bar member RB by using the image sensor DIS. The detection method may be used. For example, as disclosed in Japanese Patent No. 3077149, the alignment mark MKn on the substrate P is made into a diffraction grating mark, and two parallel beams intersecting on the diffraction grating mark are irradiated. Then, an alignment system is used in which the diffracted light generated from the diffraction grating mark by the interference fringes created by the interference of the two parallel beams is measured by homodyne measurement or heterodyne measurement, and the positional deviation of the diffraction grating mark in the pitch direction is measured. You can also.

As described above, it is formed on the substrate P according to the drawing data and the pattern data based on each position information of the alignment mark MKn on the substrate P detected by using a plurality of (two or more) alignment systems ALGn. For patterning equipment (exposure equipment, drawing equipment, printing equipment) that requires fine adjustment (correction) of the pattern position on the order of microns or submicrons, it is continuously aligned for several hours or more, and in some cases for half a day or more. The mark detection operation and the pattern drawing operation are continued without interruption. Therefore, due to changes in the temperature and humidity of the environment that may occur during that time, the influence of heat sources (motors, light sources, etc.) in the device, etc., the alignment system ALGn (detection area ADn on the substrate P) in the device. The installation position may fluctuate (drift) from the intended state.

In the alignment system ALGn described in each of the above embodiments and each modification, a beam splitter (synthetic optical member) BS1 is arranged in the optical path of the objective lens system OBL for observing the inside of the detection region ADn, and a plurality of alignment system ALGn. The reference mark RMn on the reference bar member (reference index member) RB extending along the direction in which (detection region ADn) is arranged can be appropriately observed by the alignment system ALGn via the beam splitter BS1. As a result, the amount of change from the desired state of each installation position of the plurality of alignment systems ALGn (detection area ADn) and the relative positional relationship of each installation position can be determined during the drawing operation or during the detection operation of the alignment mark MKn. It can be measured at any timing.

The sequence for measuring each fluctuation amount of the detection region ADn (imaging region DIS'of the image sensor DIS) of the alignment system ALGn and the relative positional relationship based on the reference mark RMn on the reference bar member RB is described first. Is executed as described in step 302 in FIG. 12, and is measured as the installation error information ΔCn shown in FIG. In this way, the sequence for measuring the installation error information ΔCn for each reference bar member RB of the detection region ADn (imaging region DIS'of the image sensor DIS) of the plurality of alignment systems ALGn is set as the second measurement step.

Further, the alignment system ALGn described in each of the above-described embodiments and modifications is based on the alignment mark MKn on the substrate P or the reference patterns FMa to FMh on the outer peripheral surface of the rotating drum DR as the substrate support mechanism. The reference mark RMn on the bar member RB can be simultaneously detected by the image sensor DIS. Therefore, as described in step 318 in FIG. 12, the misalignment error of each of the alignment marks MKn arranged in the width direction on the substrate P is directly set with reference to each of the plurality of reference marks RMn. Can be measured. As described above, the relative misalignment error between the reference mark RMn of the reference bar member RB and the alignment mark MKn (or the reference patterns FMa to FMh on the rotating drum DR) on the substrate P is caused by each of the plurality of alignment systems ALGn. The sequence for directly measuring the above is defined as the first measurement step.

Further, as described in step 318 in FIG. 12, each of the plurality of alignment systems ALGn is set in the image pickup region DIS'of the image pickup device DIS as shown in FIG. It is also possible to measure the misalignment error of the alignment mark MKn (or the reference patterns FMa to FMh on the rotating drum DR) on the substrate P with respect to the center point CCn). In this case, since the reference point is virtually determined in the image analysis unit 206 (see FIG. 11) that analyzes the video signal Vsg from the image sensor DIS, the image analysis unit 206 is exclusively for the alignment mark MKn on the substrate P. You only need to analyze the magnified image. Therefore, the load of the arithmetic processing for obtaining the misalignment error is reduced, and there is an advantage that one misalignment measurement of the alignment mark MKn is completed in a short time as compared with the first measurement step described above.

When such a measurement sequence is used as the third measurement step, the misalignment error of each of the alignment marks MKn measured in the third measurement step includes the detection region ADn of the alignment system ALGn (that is, the imaging region DIS). ') Installation error ΔCn with respect to the reference bar member RB is not included. Therefore, with reference to the reference mark RMn of the reference bar member RB, the amount of misalignment in the main scanning direction and the sub-scanning direction (Y direction) of the exposed region DPA on the substrate P and the amount of deformation of the substrate P in the plane ( When obtaining expansion / contraction, distortion, inclination, etc.), the misalignment error of the alignment mark MKn measured in the third measurement step is corrected by the installation error information ΔCn obtained in the second measurement step, and the error is corrected. A pattern drawing operation (step 320 in FIG. 12) is performed based on the information on the misalignment error obtained as a result of the correction.

ADn, ARn ... Detection area ALGn ... Alignment system BS1 ... Beam splitter Bsp ... Optical splitting surface EX ... Pattern forming device ILU ... Lighting system MKn ... Alignment mark OBL ... Objective lens system P ・・・ Substrate RB ・・・ Reference bar member (reference index member)
RMn ... Reference mark Un ... Drawing unit (pattern formation mechanism)

Claims

A pattern forming apparatus that forms a pattern for an electronic device in a predetermined area on a substrate that moves in the first direction.
On the surface of the substrate, there is a pattern forming region in which the dimension in the second direction orthogonal to the first direction is set longer than the dimension in the first direction, and the predetermined size of the substrate is set within the pattern forming region. A pattern forming mechanism that forms the pattern in the region,
The first substrate marks formed on the substrate at predetermined intervals along the first direction are optically formed in the first detection region set on the upstream side of the pattern formation region with respect to the first direction. The first alignment system to detect and
A second substrate mark formed on the substrate at a predetermined interval along the first direction is formed on the substrate by a predetermined distance from the first substrate mark in the second direction, and the pattern is formed in the first direction. A second alignment system that optically detects within the second detection region set on the upstream side of the region,
A portion extending in the second direction along the first alignment system and the second alignment system and corresponding to each of the first detection region and the second detection region in the second direction. A reference index member with a reference index mark formed on it,
A synthetic optical member provided in each of the first alignment system and the second alignment system and synthesized so that the light from the reference index mark passes through the optical path through which the light from the substrate mark passes.
A pattern forming device equipped with.
The pattern forming apparatus according to claim 1.
The reference index mark formed on the reference index member is separated in the second direction by a distance corresponding to a design distance between the first detection region and the second detection region in the second direction. A first reference index mark formed and detected by the first alignment system via the composite optical member and a second reference index mark detected by the second alignment system via the composite optical member. And, including, pattern forming apparatus.
The pattern forming apparatus according to claim 2.
Each of the first alignment system and the second alignment system projects illumination light toward the detection region set on the substrate and incidents light from the substrate mark appearing in the detection region. Has an objective lens system,
The synthetic optical member is arranged in an optical path between the substrate and the objective lens system, projects the illumination light toward the reference index mark of the reference index member, and emits light from the reference index mark. A pattern forming apparatus, which is a beam splitter that leads the light beam to the objective lens system.
The pattern forming apparatus according to claim 3.
The beam splitter is a parallel plate-shaped transmissive optical glass material that is arranged at a predetermined angle with respect to a plane perpendicular to the optical axis of the objective lens system.
A pattern forming apparatus in which the predetermined angle is set in the range of 5 ° to 30 °.
The pattern forming apparatus according to claim 3.
The beam splitter has a wavelength selectivity that transmits light in the first wavelength band of the illumination light projected from the objective lens system and reflects light in a second wavelength band different from the first wavelength band. A pattern forming apparatus having an optical dividing surface on which a multilayer film is formed.
The pattern forming apparatus according to claim 3.
The beam splitter has a polarization selectivity that transmits the first linearly polarized light of the illumination light projected from the objective lens system and reflects the second linearly polarized light different from the first linearly polarized light. A pattern forming apparatus having an optical dividing surface on which a multilayer film is formed.
The pattern forming apparatus according to any one of claims 1 to 6.
The reference index member extends in a bar shape in the second direction, and has a polygonal cross section perpendicular to the second direction, and extends in the second direction corresponding to one side of the polygon. A pattern forming apparatus, which is a reference bar member having a surface as a reference surface and having the first reference index mark and the second reference index mark formed on the reference surface.
The pattern forming apparatus according to claim 7.
The base material of the reference bar member is an Invar alloy, a Kovar alloy, a material obtained by mixing and sintering HfW 2 O 8 (or ZrW 2 O 8 ) and MgWO 4, and any of quartz, cordylite ceramics, and glass ceramics. A pattern forming device composed of a material having a low coefficient of thermal expansion.
The pattern forming apparatus according to claim 1 or 2.
The reference index member is provided with an illumination unit for independently illuminating the reference index mark, and the light generated from the reference index mark by the illumination unit passes through the synthetic optical member to the first alignment system. A pattern forming apparatus set to be incident on each of the second alignment systems.
The pattern forming apparatus according to claim 9.
Each of the first alignment system and the second alignment system projects illumination light toward the detection region set on the substrate and incidents reflected light from the substrate mark appearing in the detection region. Has an objective lens system
The synthetic optical member is arranged in the optical path between the substrate and the objective lens system, and incidents the reflected light from the substrate mark to guide the objective lens system, and also emits the light from the reference index mark. A pattern forming apparatus that is a beam splitter that leads to the objective lens system.
The pattern forming apparatus according to claim 10.
The reference index member extends in a bar shape in the second direction, and has a polygonal cross section perpendicular to the second direction, and extends in the second direction corresponding to one side of the polygon. A pattern forming apparatus, which is a reference bar member having a surface as a reference surface and having the first reference index mark and the second reference index mark formed on the reference surface.
The pattern forming apparatus according to claim 10.
The beam splitter is a parallel plate-shaped transmissive optical element that is tilted by a predetermined angle with respect to a plane perpendicular to the optical axis of the objective lens system.
A pattern forming apparatus in which the predetermined angle is set in the range of 5 ° to 30 °.
A pattern forming method in which a new pattern is superposed on a base pattern for an electronic device formed in a predetermined region on a substrate moving in the first direction.
Regarding the movement of the substrate in the first direction, the first of the first alignment system set on the upstream side of the pattern forming region by the pattern forming mechanism for forming the new pattern in the predetermined region of the substrate. A first mark detection step of sequentially optically detecting a plurality of first substrate marks formed on the substrate at predetermined intervals along the first direction in the detection region.
Within the second detection region of the second alignment system, which is on the upstream side of the pattern formation region and is set at a predetermined interval in the second direction orthogonal to the first direction from the first detection region. A second mark detection step of sequentially optically detecting a plurality of second substrate marks formed on the substrate at predetermined intervals along the first direction.
In designing the predetermined interval between the first detection region and the second detection region on the reference index member extending in the second direction along the first alignment system and the second alignment system. The first reference index mark of one of the two reference index marks formed at each of the positions corresponding to the distances of the above is arranged in the optical path of the first alignment system during the first mark detection step. The second reference index mark, which is the other of the two reference index marks, is detected simultaneously with the first substrate mark via the composite optical member, and the optical path of the second alignment system is detected during the second mark detection step. By simultaneously detecting the second substrate mark via the synthetic optical member arranged inside, each position of the first substrate mark and the second substrate mark with reference to the reference index mark is measured. 1 measurement process and
When the new pattern is formed in the predetermined region on the substrate by the pattern forming mechanism, it is based on the positions of the first substrate mark and the second substrate mark measured in the first measurement step. The first adjustment step of adjusting the position of the new pattern and
A pattern forming method including.
The pattern forming method according to claim 13.
The first alignment system has a first image pickup element that captures an image of the first substrate mark appearing in the first detection region or an image of the first reference index mark via the composite optical member. ,
The second alignment system has a second image pickup element that captures an image of the second substrate mark appearing in the second detection region or an image of the second reference index mark via the composite optical member. ,
The relative position error between the position of the image of the first reference index mark in the first imaging region of the first image sensor and a predetermined reference point in the first imaging region is measured, and the second By measuring the relative position error between the position of the image of the second reference index mark in the second image pickup region of the image sensor and the predetermined reference point in the second image pickup region, the reference index member can be moved. A pattern forming method further comprising a second measurement step of acquiring installation error information of each of the first alignment system and the second alignment system as a reference.
The pattern forming method according to claim 14.
The position error of the image of the first substrate mark appearing in the first imaging region of the first image sensor with respect to the reference point is measured, and the second image sensor appears in the second imaging region of the second image sensor. A pattern forming method further comprising a third measurement step of obtaining position deviation information of the predetermined region on the substrate by measuring a position error of an image of a substrate mark with respect to the reference point.
The pattern forming method according to claim 15.
In the first adjustment step, after correcting the misalignment information of the predetermined region on the substrate obtained in the third measurement step with the installation error information earned in the second measurement step, the first adjustment step is performed. The corrected misalignment information is used in place of the information on the positions of the first substrate mark and the second substrate mark measured in the first measurement step to adjust the position of the new pattern. , Pattern formation method.