JP2005184034A - Aligner and method of forming pattern using same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To avoid distortions of a mask substrate and reduce a transfer error of the pattern which is transferred onto a photosensitive substrate. <P>SOLUTION: In the aligner having a movable stage in holding the mask substrate to expose a pattern formed on the mask substrate on a photosensitive substrate by a projection optical system, the stage holds the mask substrate through an elastic member being elastically deformable according to deformation of the mask substrate. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to an exposure apparatus that is used in a lithography process when manufacturing a semiconductor device, a liquid crystal display element or the like, and transfers a circuit pattern onto a photosensitive substrate such as a semiconductor wafer or a glass plate, and further uses such an exposure apparatus. The present invention relates to a method of forming a circuit pattern on a sensitive substrate.

  At the manufacturing site of VLSI semiconductor devices and liquid crystal display devices, it is essential to perform a lithography process that exposes and transfers the circuit patterns of these devices onto a photosensitive substrate such as a semiconductor wafer or glass plate coated with a resist. is there. In this lithography process, an apparatus that projects and exposes an image of a circuit pattern formed on a mask substrate (reticle) onto a photosensitive substrate via a reduction projection optical system or an equal magnification projection optical system, a so-called stepper is used as a main device. Yes. As is well known, such a stepper is a two-dimensional structure that holds a photosensitive substrate for exposure to another shot area each time a projected image of a reticle circuit pattern is exposed to the shot area on the photosensitive substrate. It is used in a step-and-repeat system that repeats moving the moving stage by a certain amount.

  In recent years, in order to ensure higher resolution and a wide exposure field at the same time, the reticle and the photosensitive substrate are scanned one-dimensionally relative to the field of the projection optical system during shot exposure, and only the photosensitive substrate is exposed during non-exposure. A step-and-scan type projection exposure apparatus that moves in steps has also been put into practical use (pages 424 to 433 of "Optical / Laser Microlithography II" in 1989, SPIE Vol. 1088).

  In these projection exposure apparatuses, the reticle is fixedly supported on the reticle stage in the apparatus by vacuum suction so that the main surface of the reticle on which the circuit pattern is formed is positioned exactly in line with the object plane of the projection optical system. Is done. By irradiating the pattern area (usually rectangular) of the mounted reticle with exposure illumination light, an image of the pattern formed in the pattern area is projected through the projection optical system. Ordinary reticles are made by etching a light-shielding material (chrome) layer deposited on the main surface of a quartz plate into the shape of a circuit pattern, but some types of phase shift reticles are made of quartz. The shifter layer made of a transparent material formed on the main surface of the plate is made by etching according to the shape of the circuit pattern.

  Such reticles have become as large as 4 inches and 5 inches as the degree of device integration has increased and the device size has increased, and now 6-inch reticles have become standard. In an exposure apparatus for a liquid crystal device that exposes a circuit pattern by a scanning method using an equal magnification projection optical system, the mask and the photosensitive substrate (glass plate) have the same dimensions. May be used.

  By the way, mass production of 64M • D-RAM is now being started at the manufacturing site of semiconductor devices, and at the trial production level, R & D for mass production of 256M • D-RAM is being conducted, and 1G • D-RAM is being prototyped. It is done vigorously. For such mass production of devices of 256M memory and 1G memory, it is predicted that a projection exposure apparatus using an ultraviolet light source will still be used.

  On the other hand, the accuracy budget of various functions required for a projection exposure apparatus for manufacturing such a device becomes increasingly severe. In particular, an imaging accuracy required for a projection optical system, a reticle, and a photosensitive substrate (wafer). The alignment accuracy is almost limited. For this reason, it is necessary to have a design method and a manufacturing method that bring the imaging performance of the projection optical system close to an ideal one, and it is necessary to develop and improve various sensors for improving the alignment accuracy.

  Even if various performances on the projection exposure apparatus side are improved as described above, the problem of reticle (mask substrate) deformation remains as one problem. Deformation of the reticle is caused by a state of holding the reticle on the reticle stage, a change in the state of the reticle after being held once, and the like. If the reticle main surface becomes flat due to the deformation of the reticle, the projected pattern image will have distortion, curvature of field, image plane, even if the imaging characteristics of the projection optical system are ideal. An error such as tilt occurs. Therefore, the magnitude of this error affects the final image performance of the circuit pattern transferred onto the photosensitive substrate. Here, as a cause of reticle deformation, thermal expansion of the reticle due to illumination light being irradiated onto the reticle can be cited. If expansion occurs while the reticle is rigidly fixed on the reticle stage, the free expansion is constrained and the reticle is physically distorted, causing various errors as described above in the projected pattern image. To do.

  The present invention has been made in view of the problems as described above, and prevents the reticle from being deformed when the reticle is held on the reticle stage, and a projected image such as distortion aberration, field curvature, and field tilt. It is an object of the present invention to provide an exposure apparatus that does not deteriorate. Furthermore, the present invention provides a circuit pattern forming method capable of forming a circuit pattern without causing deterioration of a projected image even when the reticle held on the reticle stage undergoes thermal expansion. With the goal.

  In order to solve the above-described problems, an exposure apparatus of the present invention includes a stage (RST, RST ′) that can move while holding a mask substrate (R), and has a pattern (PA) formed on the mask substrate. An exposure apparatus that exposes a sensitive substrate (W) by a projection optical system (PL), the stage of which is an elastic member (37A to 37E, 54A to 54A) that can elastically deform the mask substrate in accordance with deformation of the mask substrate. 54C). According to this, even when the mask substrate is held and when the mask substrate expands after being held, distortion generated in the mask substrate can be suppressed.

  In this case, the exposure apparatus of the present invention measures the temperature of the mask substrate and detects the deformation amount due to thermal expansion and contraction of the mask substrate based on the measurement result (30A, 30B, 31A, 31B, 24). Furthermore, a control device (80) for adjusting the imaging characteristics of the projection optical system based on the detection result of the deformation amount detection device may be provided. Here, the control device may correct the position of the mask substrate based on the detection result of the deformation amount detection device.

  Further, another exposure apparatus according to the present invention irradiates a mask substrate (R) on which a pattern (PA) is formed with illumination light for exposure, and on the sensitive substrate (W) via the projection optical system (PL). In the exposure apparatus for transferring the pattern, a holding device (RST, RST ′) for holding the mask substrate so as to allow free expansion, and a temperature measuring device (30A, 30B) for measuring the temperature of the mask substrate. 31A, 31B), a detection device (24) for detecting an expansion component based on the measured temperature, and a control device (80) for adjusting the imaging characteristics of the projection optical system based on the calculated expansion component It is characterized by having. According to this, since the free expansion of the mask substrate is allowed, no distortion occurs due to the expansion of the mask substrate, and the expansion amount is corrected by adjusting the imaging characteristics of the projection optical system. be able to.

  In the circuit pattern forming method of the present invention, a pattern image in the pattern area is projected onto the pattern area (PA) formed on the mask substrate (R) by projecting optical light (PL). A method of forming a circuit pattern on a sensitive substrate (W) through a step of placing the mask substrate on the object plane side of the projection optical system to allow free expansion of the mask substrate; and A step of measuring a temperature of the mask substrate, a step of detecting an expansion amount from the initial state of the mask substrate generated while the sensitive substrate is exposed by the mask substrate based on the measured temperature, In order to correct a minute change due to the expansion of the mask substrate in the state of the pattern image to be exposed to the shot area of the image, the imaging of the imaging projection system according to the detected expansion amount Characterized in that it comprises a step of adjusting the resistance. According to this, even when the mask substrate is expanded, a slight change in the pattern image due to the expansion is corrected without causing distortion, and an optimum circuit pattern can be formed.

  As described above, according to the exposure apparatus of the present invention, since the mask substrate is held via the elastic member that can be elastically deformed in accordance with the deformation of the mask substrate, the distortion generated in the mask substrate can be suppressed. The error of the projected pattern image can be minimized. In addition, according to the circuit pattern forming method of the present invention, even if the mask substrate is expanded, a slight change in the pattern image due to expansion is corrected without causing distortion, so that a circuit pattern can be formed appropriately. Can do.

  A configuration of a projection exposure apparatus according to the first embodiment of the present invention will be described with reference to FIGS. 1, 2, and 3. FIG. FIG. 1 is a diagram schematically showing the overall configuration of a step-and-repeat projection exposure apparatus. In FIG. 1, the illumination optical system IL irradiates the circuit pattern region of the reticle R with illumination light for exposure with a uniform illuminance distribution. The reticle R is placed on a placement portion (contact portion) 10 that is formed to project at four locations on the reticle stage RST. Here, a specific one or two of the four placement portions 10 are used. Only the mounting portions 10 hold the reticle R by vacuum suction.

  The alignment systems 20A and 20B photoelectrically detect the reticle alignment marks formed at two positions around the reticle R, and the reticle R has a predetermined accuracy with respect to the optical axis AX of the illumination optical system IL or the projection optical system PL. Used for positioning with. The positioning is performed by the drive system 12 that translates the reticle stage RST in the XY plane perpendicular to the optical axis AX and rotates it slightly in the XY plane.

  Further, reticle stage RST is movably held on reticle stage base structure CL1 constituting a part of the column structure of the apparatus main body, and a motor and the like of drive system 12 are also mounted on base structure CL1. A beam interference portion (such as a beam splitter) of the reticle interferometer system IFR, which is a feature of the present invention, is also attached to the base structure CL1. The interferometer system IFR projects a length measuring beam BMr vertically onto a mirror surface portion formed on a part of the end surface of the reticle R, receives the reflected beam, and measures the positional change of the reticle R.

  On the other hand, the image of the circuit pattern of the reticle R is formed and projected at a reduction magnification of 1/4 or 1/5 onto the wafer W as the photosensitive substrate via the projection optical system PL disposed immediately below the reticle stage RST. The The lens barrel of the projection optical system PL is fixed to a lens base structure CL3 that constitutes a part of the column structure, and the lens base structure CL3 is provided with a reticle base structure CL1 via a plurality of support pillar structures CL2. I support it.

  The reticle interferometer system IFR shown in FIG. 1 is configured such that the reflected beam of the length measuring beam BMr interferes with the reference beam reflected by the reference mirror FRr fixed on the projection optical system PL. However, an interferometer system having a configuration in which the reference mirror is fixed to the reticle base structure CL1 side or an interferometer system incorporating the reference mirror itself may be used.

  The lens base structure CL3 is mounted on the wafer base structure CL4 on which the wafer stage WST on which the wafer W is mounted and which moves two-dimensionally along the XY plane is mounted. In this wafer stage WST, a wafer holder that vacuum-sucks the wafer W so that the surface of the wafer W coincides with the imaging surface of the projection optical system PL, and the wafer holder is moved minutely in the Z direction (optical axis AX direction). A leveling table that is slightly inclined is provided.

  The movement coordinate position of wafer stage WST in the XY plane and the minute rotation amount due to yawing are measured by wafer interferometer system IFW. This interferometer system IFW projects a laser beam from a laser light source LS onto a movable mirror MRw fixed to the leveling table of wafer stage WST and a fixed mirror FRw fixed to the bottom of projection optical system PL. The coordinate position of the wafer stage WST and the minute rotation amount (yawing amount) are measured by making the reflected beams from the mirrors MRw and FRw interfere.

  On the leveling table of wafer stage WST, various alignment systems, focus sensors, leveling sensor calibration and baseline measurement (measurement of positional relationship between projection center of reticle R pattern center and detection center of each alignment system) ) And a reference plate FM to be used. On the surface of the reference plate FM, a reference mark that can be detected by the alignment systems 20A and 20B is formed together with the mark on the reticle R under illumination light having an exposure wavelength.

  The baseline measurement method using the reference plate FM is disclosed in, for example, Japanese Patent Application Laid-Open Nos. 4-45512 and 5-21314, and the focus calibration method using the reference plate FM is disclosed. For example, since it is disclosed in Japanese Patent Publication No. 6-16483, detailed description thereof is omitted here. FIG. 2 is a perspective view showing the state in the vicinity of reticle stage RST in FIG. 1. Here, reticle R is shown as being correctly aligned with optical axis AX of projection optical system PL. At the center of the reticle stage RST, the projection optical path between the pattern area PA formed in the rectangular shape of the reticle R and the two marks M1 and M2 formed on both sides in the X direction across the pattern area PA is not shielded. Such an opening 11 is formed. At four locations around the opening 11, mounting portions 10A, 10B, 10C, and 10D (however, 10C and 10D are shaded by the reticle R in FIG. 2) are formed at a predetermined height. Thus, the vicinity of the four corners of the back surface of the reticle R (the main surface on which the pattern is formed) is supported.

  Specifically, the drive system 12 of the reticle stage RST shown in FIG. 1 includes three drive units 12A, 12B, and 12C as shown in FIG. The drive unit 12A slightly moves the stage RST in the X direction, and the two drive units 12B and 12C cooperate with each other to slightly move the stage RST in the Y direction and slightly rotate it. These drive units 12A, 12B, and 12C are configured as actuators that linearly press corresponding portions of the reticle stage RST, and there are reticles between each of the several portions of the reticle stage RST and the reticle base structure CL1. A spring 13 is provided to bias the stage RST in the direction of each actuator.

  As shown in FIG. 2, reflection portions SX, SY, and Sθ for receiving a length measurement beam from the interferometer system IFR for the reticle are formed on the surfaces of the two end surfaces of the reticle R that are orthogonal to each other. Has been. These reflecting portions are formed by polishing a reflective metal material such as aluminum or chrome, or a dielectric thin film having a high reflectivity with respect to the wavelength of the measurement beam after polishing the end face of the reticle R. The reflection portions SX, SY, Sθ are formed only on a part of each end face of the reticle R in FIG. 2, but one end face extending in the X direction and one end face extending in the Y direction May be a reflective surface.

  The reticle interferometer system IFR shown in FIG. 1 specifically includes an X-direction interferometer IFRX including a reflection mirror BR1 that projects the measurement beam BMrx toward the reflection unit SX in the X direction as shown in FIG. A Y-direction interferometer IFRY including a reflection mirror BR2 that projects the measurement beam BMry toward the reflection unit SY in the Y direction; and a reflection mirror BR3 that projects the measurement beam BMrθ in the Y direction toward the reflection unit Sθ. It is composed of a Y-direction interferometer IFRθ.

  Here, the length measurement beam BMrx reaching the reflection part SX is arranged to be perpendicular to the optical axis AX of the projection optical system PL, and the length measurement beam BMry reaching the reflection part SY and the length measurement beam BMrθ reaching the reflection part Sθ, Are arranged in parallel in the X direction at a predetermined interval. Therefore, the position change of the reticle R in the X direction is measured by the measured value of the interferometer IFRX, and the position change of the reticle R in the Y direction is calculated by calculating the average of the measured values of the two interferometers IFRY and IFRθ. The position change in the rotation (θ) direction of the reticle R is measured by calculating the difference between the measured values of the two interferometers IFRY and IFRθ.

  The measurement resolution of the interferometers IFRX, IFRY, and IFRθ can detect the amount of expansion determined depending on the linear expansion coefficient of the reticle R, the assumed temperature change range of the reticle R, and the size of the reticle R (here, 6 inches). Set to degree. Further, when the cross sections of the measurement beams BMrx, BMry, and BMrθ of each interferometer are viewed on the reflection portion of each end face of the reticle R, they are shaped into an elliptical shape or a slit shape that is flattened in the extending direction of the end face. Further, the cross section of each reference beam projected onto the reference mirror FRr fixed to the upper part of the projection optical system PL is also shaped into an elliptical shape or a slit shape that is flat in the same direction as the cross sectional shape of each length measuring beam.

  The cross sections of the length measurement beam and the reference beam are thus made elliptical or slit-like because the traveling direction of the reflected beam on the reticle end surface of each length measurement beam is minute due to the minute rotation of the reticle R in the XY plane. This is to prevent the deviation from the interference state due to the deflection within the range of rotation of the reticle R. FIG. 3 is a plan view of the upper surface portion of the reticle stage RST shown in FIG. 2, and the placement portions 10A, 10B, 10C, and 10D that support the reticle R at four positions include a pattern area PA and a mark. It is provided at the four corners of the opening 11 so as not to shield the projection light paths of M1 and M2. Among these, the suction surface (abutment surface) of the mounting portion 10C that is separated from any of the reflection portions SX, SY, and Sθ of the reticle R irradiated with each of the measurement beams BMrx, BMry, and BMrθ of the interferometer system for the reticle ) Is formed larger than the suction surfaces of the other three mounting portions 10A, 10B, and 10D, and the suction force during vacuum suction is increased.

  Also, as shown in FIG. 3, there are ports 15A, 15B, 15C, 15D for connecting the suction surfaces of the mounting parts 10A, 10B, 10C, 10D to a reduced pressure source (vacuum source) around the reticle stage RST. Is provided. Accordingly, it is possible to firmly hold the four corners of the reticle R on the stage RST by supplying a predetermined reduced pressure to each of the port portions 15A to 15D. However, in the present embodiment, while at least a plurality of wafers W are continuously subjected to the exposure process, only the mounting unit 10C strongly sucks the reticle R under reduced pressure, and the other mounting units simply hold the reticle R under its own weight. Adsorption is released to support.

  In this way, when the reticle R is strongly sucked under reduced pressure only by the mounting portion 10C, the expansion of the reticle R due to the temperature change is in the direction of the reflecting portions SX, SY, Sθ formed on the two end surfaces with the mounting portion 10C as a base point. It is generated freely, and the stress generated in the reticle R along with the expansion is prevented, resulting in a linear expansion substantially along the XY plane. Therefore, the reflecting portions SX, SY, Sθ are slightly displaced in the X direction and the Y direction according to the free expansion amount of the reticle R, and the minute displacements are transmitted through the measurement beams BMrx, BMry, BMrθ. It is measured by the interferometer systems IFRX, IFRY, and IFRθ.

  By the way, there is a case where the reticle R slightly deviates from the stage RST against the strong adsorption force by the mounting portion 10C. In this case, the reticle interferometer systems IFRX, IFRY, and IFRθ output measurement values including the shift amount and the expansion amount of the reticle R. For this reason, based on only the measurement values obtained from the interferometer systems IFRX, IFRY, and IFRθ for the reticle, the displacement amount (displacement in the X and Y directions) and the minute rotation amount of the center point CC of the reticle R from the initial state. It is not possible to accurately determine whether the amount of positional deviation or the amount of rotation is due to the expansion of the reticle R, whether the reticle R is displaced on the stage RST, or the cause of both. .

  Therefore, the positional deviation amount and the rotation amount value calculated based only on the measurement value from the reticle interferometer system are not reliable as they are, and based on that, the drive unit 12A, 12B, 12C of the reticle stage RST is immediately set. Feedback control is not possible. Therefore, in order to accurately specify the displacement error from the initial position of the center point CC of the reticle R and the rotation error of the reticle R, a temperature measurement system for measuring the temperature of the reticle R with high resolution is provided in this embodiment. I made it. When a thermocouple type sensor is used as the temperature measurement system, it is necessary to provide a mechanism for bringing the sensor into contact with a plurality of locations avoiding the pattern area PA of the reticle R and the formation areas of the marks M1 and M2. When a non-contact type thermosensor (such as an IR-CCD camera) is used, a temperature change (and temperature distribution) on almost the entire surface of the reticle R is detected through the condenser lens system in the illumination optical system IL shown in FIG. It is necessary to provide an optical system.

  Then, the linear expansion coefficient of the transparent flat plate (quartz plate) of the reticle R obtained in advance, the amount of change from the measured reference temperature of the reticle R (for example, the set temperature at the time of reticle manufacture), and the reticle by the mounting unit 10C Based on the relationship between the support point on R and the distance between each reflection part SX, SY, Sθ, the positional change amounts Δmx, Δmy, Δmθ of the reflection parts SX, SY, Sθ due to only the expansion component of the reticle R are predicted. Then, based on the calculation result, the position displacement amount (ΔXt, ΔYt) and the reticle rotation amount ΔRt due to only the expansion component of the center point CC of the reticle R are determined.

Next, the amount of change ΔFx, ΔFy, ΔFθ from the initial position of each of the reflectors SX, SY, Sθ measured by the reticle interferometer system IFRX, IFRY, IFRθ is read, and the predicted amounts of change Δmx, Δmy, Differences ΔMx, ΔMy, and ΔMθ from Δmθ are calculated as follows.
ΔMx = ΔFx−Δmx (X direction component)
ΔMy = ΔFy−Δmy (Y direction component)
ΔMθ = ΔFθ−Δmθ (Y direction component)
If each of the deviations ΔMx, ΔMy, and ΔMθ calculated in this way is approximated to zero within an allowable range, the reticle R is not displaced on the stage RST, and is predicted and calculated only by the expansion component. The drive units 12A, 12B, and 12C of the reticle stage RST may be feedback controlled so that the position displacement amount (ΔXt, ΔYt) and reticle rotation amount ΔRt are corrected.

  On the other hand, when each of the calculated deviations ΔMx, ΔMy, ΔMθ has an eigenvalue greater than or equal to the allowable range, it indicates that each of the deviations ΔMx, ΔMy, ΔMθ is caused by the positional deviation of the reticle R. Therefore, based on the deviations ΔMx, ΔMy, and ΔMθ, the position displacement amount (ΔXs, ΔYs) of the center point CC due to the displacement of the reticle R and the reticle rotation amount ΔRs are determined as follows. However, Lk is an interval in the X direction between two length measuring beams BMry and BMrRθ of the reticle interferometer as shown in FIG.

ΔXs = ΔMx
ΔYs = (ΔMy + ΔMθ) / 2
ΔRs = (ΔMy−ΔMθ) / Lk
Further, the position displacement amount (ΔXt, ΔYt) or reticle rotation amount ΔRt obtained only from the expansion component obtained previously is added to the position displacement amount (ΔXs, ΔYs) or reticle rotation amount ΔRs due to only the displacement component of the reticle R. Thus, the positional deviation amount (ΔXs + ΔXt, ΔYs + ΔYt) and the rotation amount (ΔRs + ΔRt) of the center point CC of the reticle R to be finally corrected are determined. Then, the drive units 12A, 12B, and 12C of the reticle stage RST are adjusted to the measured values of the interferometers IFRX, IFRY, and IFRθ so that the determined positional deviation amount of the center point CC and the rotation amount of the reticle R are corrected. Based on feedback control.

  FIG. 4 shows a schematic configuration of a control system for executing various calculations and controls as described above. 4, the same members as those in FIGS. 2 and 3 are given the same reference numerals. First, as shown in FIG. 4, each of the interferometer systems IFRX, IFRY, and IFRθ for the reticle includes a reflected beam of each length measuring beam at the reflecting portions SX, SY, and Sθ and a reflected beam of the reference beam at the reference mirror. Receivers 21X, 21Y, and 21θ for photoelectrically detecting the interference beam are provided.

  Signals from each of the receivers 21X, 21Y, and 21θ are input to the counter circuit units 22X, 22Y, and 22θ, and digital values corresponding to the positions of the reflection units SX, SY, and Sθ in the measurement direction and position changes are output in real time. To do. These digital values are input to a central processing circuit (CPU) 24 that executes the various calculations and controls described above. The CPU 24 calculates a positional deviation error and a rotation error of the center point CC of the reticle R determined as described above, and outputs control information for correcting the error to the interface circuit 26. The interface circuit 26 outputs optimal control command values to the servo circuits 28A, 28B, and 28C that drive the motors of the three drive units 12A, 12B, and 12C.

  4 is provided with temperature sensors 30A and 30B that measure temperature changes at a plurality of positions on the reticle R in a contact manner, and signals from the sensors 30A and 30B are transmitted by the temperature measurement circuits 31A and 31B. After being converted to a digital value, it is input to the CPU 24. The CPU 24 reads out the linear expansion coefficient of the reticle R stored in advance in the memory based on the temperature information and information indicating the positional relationship between the reflecting portions SX, SY, Sθ on the reticle R, and sets each reflecting portion. Position change amounts Δmx, Δmy, Δmθ due to expansion in the measurement direction of SX, SY, Sθ are predicted and calculated.

  In the above embodiment, the reticle R is adsorbed by only one mounting portion 10C on the reticle stage RST. However, this is not necessarily required, and may be in a range that does not hinder the free expansion of the reticle R. For example, the other three placement units 10A, 10B, and 10D may be loosely adsorbed under reduced pressure. Further, all of the four placement portions 10A, 10B, 10C, and 10D may be controlled so as to be loosely and uniformly sucked under reduced pressure within a range that does not hinder the free expansion of the reticle R.

  As described above, according to the present embodiment, the reticle stage RST that supports the reticle R adsorbs and fixes the reticle R with the mounting portion 10C at one corner or loosely adsorbs under reduced pressure, so that the reticle R changes in temperature. Therefore, the reticle R can be freely expanded, unnecessary stress is not generated on the reticle R, and deformation of the pattern surface of the reticle R is prevented. For this reason, it is possible to prevent the quality of the projected image from gradually deteriorating while the exposure process is continuously performed.

  Next, a second embodiment of the present invention will be described with reference to FIG. FIG. 5 is a perspective view showing a structural modification of the reticle stage RST shown in FIG. 1, and the four corners of the reticle R are four mounting portions 10A, 10B, 10C, 10D similar to FIG. Supported on top of. However, in this embodiment, each mounting part simply supports the reticle R with its own weight, and vacuum suction is not performed at all or loosely.

  In the present embodiment, the movable pressing members 36A, 36B, 36C, 36D, and 36E that weakly press the respective end surfaces of the four sides of the reticle R in a state where the reticle R is supported on each mounting portion are arranged around the stage RST. Has been placed. The movable pressing members 36A to 36E are set in a state of opening outward as the movable pressing member 36E in FIG. 5 when the reticle R is conveyed onto the mounting portions 10A to 10D. When the reticle R is placed on the placement portions 10A to 10D with mechanical pre-alignment accuracy (for example, ± 1 mm or less), the five movable pressing members 36A to 36E stand up vertically to correspond to the reticle R. Is pressed in the direction of the reticle center CC.

  As a result, the reticle R is accurately positioned on the stage RST. At this time, the positioning accuracy in the rotation direction of the reticle R is such that each length measuring beam BMrx, BMry, BMrθ of the reticle interferometer system is accurately reflected by the reflecting portions SX, SY, Sθ of the corresponding reticle end surface to perform interference measurement. Is set to the extent that can be surely.

  When the positioning of the reticle R is completed in this way, each of the elastic contact portions 37A, 37B, 37C, 37D, and 37E provided at the portions facing the reticle end surfaces of the movable pressing members 36A to 36E come into loose contact with the reticle end surfaces. In addition, the movable pressing members 36A to 36E are stopped in a state where they are slightly opened outward. In this state, the reticle R is accurately positioned, and when the reticle R subsequently expands, the end surface of the reticle R slightly deforms the elastic contact portions 37A to 37E outward.

  When the reticle R is set on the stage RST as described above, the counter circuit units 22X, 22Y, and 22θ shown in FIG. 4 are reset to the initial values, and the CPU 24, interface circuit 26, servo circuits 28A, 28B, Control of the drive units 12A, 12B, and 12C is immediately started by 28C. Then, the reference mark of the reference plate FM on the wafer stage WST shown in FIG. 1 is positioned in the field of the projection optical system PL, and the marks M1 and M2 on the reticle R and the reference plate FM are aligned by the alignment systems 20A and 20B. Are detected at the same time to measure the relative displacement error amount and the rotation error amount, and give a command to the CPU 24 so that these errors are corrected.

  As a result, reticle stage RST is slightly moved, and drive units 12A, 12B, and 12C are stopped in a state where precise alignment between each of marks M1 and M2 on reticle R and the reference mark on reference plate FM is achieved. The measured values of the counter circuit units 22X, 22Y, and 22θ at that time are read and stored as initial values. This initial value represents the position of each reflection part SX, SY, Sθ before the reticle R is expanded.

  Next, a third embodiment of the present invention will be described with reference to FIG. In this embodiment, the reticle R in which the reflection portions are uniformly formed on the respective end surfaces of the four sides is used, and the expansion amount due to the temperature change of the reticle R is directly measured by a dedicated interferometer system. The interferometer system is configured to directly measure the distance between the sides of the reticle R that are parallel to each other, that is, the dimensional change of the reticle R in the X or Y direction.

One interferometer system for measuring an expansion amount includes a beam splitter BS1 that divides a laser beam B0 into two beams Bx and By, a mirror MRa that reflects the beam By in the Y direction, and a beam By from the mirror MRa. Is split into a transmitted beam directed to the end face extending in the X direction of the reticle R and a beam directed to the upper mirror MR2, and a beam splitter BSa constituting an interference portion, and a main surface of the reticle R from the mirror MR2 A corner mirror MR4 that receives a beam BTy traveling in parallel in the Y direction on the opposite side of the reticle R and reflects it vertically toward the end surface 39A of the reticle R, and a beam reflected by the end surface 39A via the corner mirror MR4 and the mirror MR2. Reflected by the reflected beam returning to the splitter BSa and the end face of the reticle R on the beam splitter BSa side. Constituted by the receiver 40Y for receiving the interference beam and the reflected beam returned to over beam splitter BSa.

  Information measured by the receiver 40Y corresponds to a change in dimension of the reticle R in the Y direction. On the other hand, the other interferometer system for measuring the amount of expansion is similarly configured, and the beam Bx is incident on the end face extending in the Y direction of the reticle R via the mirror MRb and the beam splitter BSb, and the mirror MR1 and the corner mirror. Reflected beams from the respective end surfaces that are incident on the opposite end surface 39B of the reticle R through the MR3 and interfere with the beam splitter BSb are received by the receiver 40X. Accordingly, the information measured by the receiver 40X corresponds to the change in the dimension of the reticle R in the X direction.

  In the interferometer system as described above, the beam splitter BS1, the mirrors MRa, MRb, MR1 to MR4, the beam splitters BSa, BSb as the interference units, and the receivers 40X, 40Y are basically shown in FIG. 3 and FIG. It is built on a reticle stage RST. However, when the beams Bx and By incident on the beam splitters BSa and BSb as the interference units are guided by the single mode optical fiber and the interference beams from the beam splitters BSa and BSb are guided by the optical fiber, the mirrors MRa and MRb MR1-MR4, beam splitters BSa, BSb can be mounted on reticle stage RST, and other members such as receivers 40X, 40Y can be mounted on reticle base structure CL1 in FIG.

  The reticle R shown in FIG. 6 is supported on the mounting portions 10A to 10D of the reticle stage RST as shown in FIGS. 3 and 5 described above so that free expansion is not restrained. Then, after the reticle R is positioned on the stage RST using the alignment systems 20A and 20B and the reference plate FM in FIG. 1, the counter circuit that converts each measurement information from the receivers 40X and 40Y to a digital value is reset. To start counting.

  Thereafter, if the reticle R does not expand, the count value of the counter circuit does not change, but if the expansion occurs, the count value of the counter circuit changes from the value at the time of resetting. And the amount of expansion in the Y direction are immediately known. Therefore, based on the read change value, it is possible to accurately calculate the positional shift due to the expansion component of the center point CC of the reticle R.

  Further, the projection exposure apparatus shown in FIG. 1 detects the mark on the wafer W by being fixed to the outside of the TTL alignment system or the projection optical system PL for detecting the mark on the wafer W only through the projection optical system PL. When an off-axis type wafer alignment system is provided (for example, Japanese Patent Laid-Open No. 5-21314), the mark detection center point and the reticle R of each of the TTL alignment system and the off-axis alignment system after the reticle R is mounted. It is necessary to accurately measure the relative distance (baseline) between the center CC and the projection point using the reference plate FM on the wafer stage WST.

  The baseline measurement operation may be performed not only immediately after the single reticle R is mounted on the exposure apparatus but also at appropriate timings during continuous exposure processing. In this case, the value counted by the counter circuit via the receivers 40X and 40Y during the baseline measurement operation is read by the CPU and stored as an initial value. Or you may make it reset the counter circuit connected with each receiver 40X and 40Y at the time of the operation | movement of a baseline measurement.

  According to the interferometer system for measuring the expansion amount shown in FIG. 6 above, since the expansion amount in the X direction and the expansion amount in the Y direction of the reticle R are directly measured, the interferometer systems IFRX, IFRY, IFRθ It is possible to precisely measure the distance change from each end face 39A, 39B irradiated with each length measuring beam BMrx, BMry, BMrθ to the center CC of the reticle R. Therefore, since the position change of the projection point of the center CC of the reticle R can be accurately determined in real time, the alignment accuracy between the reticle R and the wafer W at the time of exposure can be precisely managed.

  Next, a fourth embodiment of the present invention will be described with reference to FIG. In this embodiment, unlike the previous embodiments, the reticle stage RST ′ shown in FIG. 7 is fixed on the reticle column structure CL1 in FIG. 1, or only when the reticle is changed, the exposure position and the transfer position are set. Configured to reciprocate between. The reticle R is mounted on four mounting portions 10A, 10B, 10C, 10D of the stage RST ′, and three linear actuators (piezo element, small linear motor, voice coil motor, etc.) 50A, 50B, 50C positions reticle R by directly pressing the corresponding end face of reticle R. In this embodiment, it is assumed that a reticle R in which all four end faces are polished to form a reflection portion is used.

  When the reticle R is mechanically pre-aligned and placed on the placement portions 10A to 10D, the pressing element 52A of the actuator 50A is moved in the X direction toward the end surface 39A of the reticle R, and the actuators 50B and 50C are moved. The pressing elements 52B and 52C are moved in the Y direction toward the end surface 39B of the reticle R. A steel ball or a synthetic resin chip that minimizes the friction coefficient is provided at a contact portion of each of the pressing elements 52A, 52B, and 52C with the end surface of the reticle R.

  When the two end faces 39A, 39B are pressed by the pressing elements 52A, 52B, 52C, the reticle R is slightly moved in the X and Y directions, and the opposite end faces 39C, 39D are elastic contactors 54A, 54B, 54C. Abut. The elastic contact 54A is configured to make point contact with the central portion of the end surface 39D of the reticle R and to be elastically displaced in the Y direction within a predetermined range (for example, several mm). 54C is configured to make point contact with both side portions of the end surface 39C of the reticle R and to be elastically displaced in the X direction within a predetermined range (for example, several mm). The elastic force of these elastic abutments 54A, 54B, 54C is set to such an extent that the free expansion of the reticle R is not hindered.

  Thus, when the actuator R is finely moved in the X and Y directions by driving the actuators 50A, 50B, and 50C, and the elastic abutments 54A, 54B, and 54C are elastically displaced to the middle of their strokes, the reticle R is mechanically moved. In other words, the alignment is almost correct. Subsequently, using the alignment systems 20A and 20B shown in FIG. 1 and the mark on the reference plate FM on the wafer stage WST, the marks M1 and M2 on the reticle R are more precisely aligned with the reference plate FM. Then, the actuators 50A, 50B, and 50C are finely driven to precisely align the center CC of the reticle R with respect to the coordinate system on the wafer stage WST side.

  Also in this embodiment, information on the amount of expansion of the reticle R is similarly given by the temperature measurement method as shown in FIG. 4 or the direct measurement method by the interferometer system as shown in FIG. The displacement and the rotation error of the reticle R are corrected together with the positional deviation and rotational deviation of the reticle R in the XY directions with respect to the mounting portions 10A to 10D. Of these corrections, corrections in the X and Y directions may be performed not only on the reticle side actuators 50A, 50B and 50C but also on the wafer stage WST.

  As described above, in the present embodiment, since the driving force is directly applied to the end surface of the reticle R, the measurement surfaces by the measurement beams BMrx, BMry, and BMrθ of the interferometer systems IFRX, IFRY, and IFRθ for the reticle The surfaces including the drive points of the plurality of actuators substantially coincide with each other, and the control accuracy of the servo control system is improved without being affected by the slip between the mounting portions 10A to 10D and the reticle R.

  Next, a fifth embodiment of the present invention will be described with reference to FIG. FIG. 8 shows a modified example of the fiber interferometer system for directly measuring the expansion amount of the reticle R. In this embodiment, a reticle R in which an optical dividing portion is formed by a dielectric thin film on at least one side end face is used. Is called. In FIG. 8, members having the same functions as those of the previous embodiments are denoted by the same reference numerals.

  As shown in FIG. 8, the reticle R is supported on the four placement portions 10A to 10D of the reticle stage RST. Here, as in the first embodiment, any one placement portion holds the reticle R as in the first embodiment. Adsorbed under reduced pressure. Of the end face of the reticle R, a reflection part ST is formed together with the reflection parts SY and Sθ on the end face 39A on which the measurement beams BMry and BMrθ of the reticle interferometer systems IFRY and IFRθ are projected. The reflecting portion ST is formed on the surface of the end surface 39A that has been subjected to planar polishing, but is used here as a back reflecting mirror. Therefore, a reflective layer may be deposited on the entire end surface 39A.

  A partial reflection portion SS made of a dielectric thin film is formed on the end surface 39C opposite to the end surface 39A of the reticle R, and a length measuring beam BMt for measuring an expansion amount is mainly supplied from the partial reflection portion SS to the inside of the reticle R. Incident parallel to the surface. This length measurement beam BMt includes a single-mode optical fiber 60 that receives a linearly polarized beam B0 from a laser light source, a support fitting 62 that fixes the exit end of the optical fiber 60 to the reticle stage RST, and a reticle stage RST. And is projected substantially perpendicularly onto the partial reflection portion SS of the end surface 39C of the reticle R via the beam splitter 64 that functions as an interference portion. Although not shown in FIG. 8, a quarter wavelength plate is disposed between the beam splitter 64 and the end face 39C.

  Therefore, the beam projected on the partial reflection unit SS becomes circularly polarized light, and a part of the beam reflected by the partial reflection unit SS becomes linearly polarized light orthogonal to the original beam B0 by the ¼ wavelength plate. Returning to 64, the light is reflected and incident on the optical fiber 66 for light reception, and is guided to the receiver (photoelectric detection element). On the other hand, a part of the beam transmitted through the partial reflection portion SS becomes a measurement beam BMt and reaches the reflection portion ST on the opposite end face 39A perpendicularly, and the beam reflected here has the same optical path as the measurement beam BMt. Returning to the beam splitter 64 via the partial reflector SS and the quarter-wave plate. The reflected beam is reflected by the beam splitter 64 and enters the optical fiber 66.

  For this reason, the reflected beam from the partial reflector SS and the reflected beam from the reflector ST on the end face 39A interfere with each other and are received by the receiver as an interference beam BF. Accordingly, the receiver measures the change in the interval in the Y direction between the partial reflection portion SS of the end surface 39C of the reticle R and the reflection portion ST of the end surface 39A with a resolution of 1/100 μm or less. That is, the interferometer system including the beam splitter 64 and the receiver directly measures the expansion amount of the reticle R in the Y direction.

  In the case of the present embodiment, it is necessary to form a thin film layer (multilayer film) serving as the partial reflection portion SS on the end surface 39C of the reticle R, preferably with a reflectance of 33% and a transmittance of 67%. 8, the beam B0 incident on the beam splitter 64 is supplied from the optical fiber 60. A. (Numerical aperture) is desirably made as small as possible to maintain the parallelism of the beam BMt. Further, the beam B0 may be incident on the beam splitter 64 via a reflecting mirror fixed to the reticle base structure CL1 (see FIG. 1) without using the optical fiber 60.

  Next, a projection exposure apparatus according to a sixth embodiment of the present invention is described with reference to FIG. In FIG. 9, the beam splitter 64 that constitutes the interferometer system for measuring the expansion amount shown in FIG. 8 is fixed on the reticle stage RST, and each of the reflection part ST of the end face 39A and the partial reflection part SS of the end face 39C. The beam reflected by is received by the receiver 68 through the optical fiber 66. A processing unit 80 including a counter circuit and a CPU inputs a signal from the receiver 68 and calculates the expansion amount of the reticle R from the initial state in real time. In the present embodiment, a minute dimensional error (appears as a magnification error) on the wafer W of the circuit pattern image generated by the expansion of the reticle R is corrected using the measured expansion amount data.

  For this purpose, the field lens G1 disposed in the telecentric part of the projection optical system PL is structured to translate or tilt in the direction of the optical axis AX, and a plurality of piezo elements 70A and 70B that perform these movements. Is provided. Then, the drive control circuit 72 controls the drive amount of each of the piezo elements 70A and 70B, so that the imaging magnification and distortion of the projection optical system PL are minutely changed.

  Alternatively, an appropriate air gap (air lens) 73 in the projection optical system PL is sealed, and the internal pressure is adjusted by the pressure control system 74, so that the imaging magnification of the projection optical system PL is slightly changed. . Further, in order to correct the position change in the optical axis AX direction of the best imaging plane, which is generated by adjusting the position of the field lens G1 and the internal pressure of the air lens 73, the imaging light is projected obliquely onto the surface of the wafer W. Then, a plurality of detection systems 75 that receive the reflected light FL and detect the position of the wafer W in the optical axis direction and a plurality of holders WH that attract the wafer W based on the detected position of the wafer W are displaced in the Z direction. An oblique incidence autofocus system is provided which includes an actuator 77 and a drive control system 76 that drives the actuator 77.

  In the above configuration, the processing unit 80 expands or reduces the ratio (ppm) of the circuit pattern area on the reticle R based on the center CC based on the calculated amount of expansion of the reticle R from the initial state. Correction information is output to at least one of the drive control circuit 72 and the pressure control system 74 so that correction corresponding to the ratio is performed. In response to this, the position correction of the field lens G1 or the internal pressure correction of the air lens 73 is executed, and the dimensional error of the projected image on the wafer W is corrected.

  A drive control circuit 72 for controlling the position of the field lens G1, a pressure control system 74 for controlling the internal pressure of the air lens 73, or a drive control system 76 functioning as an oblique incidence autofocus system is disclosed in, for example, Japanese Patent Laid-Open No. 60-78454. As disclosed in Japanese Laid-Open Patent Publication No. 62-2229838, fluctuations in imaging characteristics due to changes in atmospheric pressure of the projection optical system PL, fluctuations in imaging characteristics due to partial absorption of illumination energy during exposure, Alternatively, it also has a function for correcting fluctuations in imaging characteristics in accordance with changes in the state of illumination light from the illumination optical system that illuminates the reticle.

  Therefore, if the amount of expansion of the reticle R from the initial state is directly measured every moment by the interferometer system as in this embodiment, the apparent magnification change due to the expansion of the reticle R can be corrected in real time. It becomes possible to correct all of the various imaging errors that occur in the imaging optical path (from the reticle pattern to the wafer surface) of the projection exposure apparatus.

  By the way, in each of the embodiments described above, the main purpose is to measure the expansion due to thermal accumulation of the reticle (mask substrate) and correct the alignment position, the projection magnification, and the like caused by the expansion, but on the reticle stage RST. If a duct that blows precisely temperature-controlled air at a predetermined flow rate is arranged toward the reticle R placed on the reticle R, and the temperature rise due to heat accumulation of the reticle R can be sufficiently reduced, the temperature of the reticle R It is also possible to omit a temperature sensor that measures the interferometer system and the like that directly measure the amount of expansion.

  However, even in such a case, the amount of displacement in the X, Y, and θ directions of the reticle R is measured by the interferometer systems IFRX, IFRY, and IFRθ using the reflecting portion (mirror surface portion) formed on the end surface of the reticle R, and the measurement is performed. It is desirable to finely move the reticle R based on the information. That is, the object to be actually positioned is the reticle R, and in the conventional method in which the reflector for the interferometer is fixed to the reticle stage RST for minute movement, the minute amount of the reticle R mounted on the reticle stage RST is small. This is because no slip (position shift) can be detected.

  By the way, in reticle R (mask substrate) used in each embodiment of the present invention, at least a part of its end face is mirror-polished, and a reflective layer made of a reflective material is formed thereon by vapor deposition. Such a reflective layer is preferably formed at a stage prior to drawing a circuit pattern on a quartz substrate which is a base material of the reticle R. In a normal reticle R manufacturing process, optical polishing is performed so that the parallelism of both sides of a quartz substrate (6 inches, 5 mm thickness) as a base material is within an allowable value, and then one side of the polished quartz substrate Low reflection chromium is deposited on the entire surface, and a resist for electron beam exposure (EB exposure) is applied thereon. Then, the quartz substrate coated with the resist is placed in a pre-aligned state on the sample stage of the EB exposure apparatus, and drawing of a circuit pattern, a mark pattern, and the like is started.

  The exposed quartz substrate is etched by masking the portion of the resist image remaining on the chromium layer by the development process. By this etching, the portion of the chromium layer that should be transparent in the circuit pattern is removed to form the reticle R. Therefore, in order to form a reflection part and a partial reflection part on the end face of the reticle R, in such a manufacturing process, all four end faces are polished together at the stage of polishing the quartz substrate as a base material, In the step of depositing the chromium layer on the entire surface of the base material of the quartz substrate, the chromium layer (or dielectric multilayer film) may be deposited together on all or some of the end faces of the four sides.

  Then, on the sample stage of the apparatus that performs EB exposure on the resist on the quartz substrate thus prepared, the quartz substrate comes into contact with two orthogonal end faces (for example, 39A and 39B in FIGS. 6 and 8) of the quartz substrate. A reference pin or the like for positioning the substrate is provided, and the quartz substrate is placed on the sample stage while being positioned with respect to the reference pin, and EB exposure is performed. In this case, the drawn circuit pattern and alignment mark are arranged with reference to the two end faces of the quartz substrate.

  For this reason, the arrangement of the circuit pattern is precisely defined from the two end faces, and the reference of the pattern arrangement on the completed reticle R is based on the reticle interferometer system of the projection exposure apparatus described in the previous embodiments. It coincides with the measurement point. Accordingly, by matching the reference, it is possible to suppress a decrease in overlay accuracy with the wafer W due to a pattern drawing error (placement error) on the reticle R.

  A movable mirror for a length measuring system using a laser interferometer is fixed to the peripheral edge of the sample stage of the EB exposure apparatus. Therefore, a seventh embodiment in which the configuration around the sample stage of the EB exposure apparatus is improved as shown in FIG. As shown in FIG. 10, interferometers EIX and EY for projecting measurement beams Bex and Bey are provided on the end faces 39A and 39B of the quartz substrate QP placed on the sample stage 90, and the interferometers EIX and EYY are provided. It is desirable to control the movement of the sample stage 90 on the basis of the reflection surfaces of the end faces 39A and 39B of the quartz substrate QP in response to the measurement result of.

  However, the end faces 39A and 39B of the quartz substrate QP are generally not so flat in the extending direction. Therefore, the position control of the sample stage 90 is performed by interferometer systems SIX and SIY using the same moving mirrors MRwx and MRwy fixed on the sample stage 90, and the interferometers EIX and EY provided separately are respectively Used to measure the flatness of the end faces 39A, 39B.

  In FIG. 10, AX ′ is the optical axis of the electron lens system of the EB exposure apparatus, PBF is a drawing electron beam (or charged particle beam), and SA is one shot that can be drawn by deflection of the beam PBF. It is a range. At this time, the sample stage 90 is linearly moved, for example, in the X direction based on the coordinate values from the interferometer systems SIX and SIY, and the measurement values of the end face 39A measured by the interferometer EIY are sequentially sampled during the movement. Thus, the flatness of the end face 39A and the parallelism with the X axis are obtained. The flatness and parallelism of the end face 39B can also be obtained by sequentially reading the measurement values of the interferometer EIX while linearly moving the sample stage 90 in the Y direction based on the coordinate values from the interferometer systems SIX and SIY. .

  When the pattern is actually drawn, the drawing position error caused by the error between the flatness and the parallelism of the obtained end faces 39A and 39B is used as the feed position of the sample stage 90 and the deflection of the drawing beam PBF. It may be finely corrected by correcting the position. In this way, the circuit pattern is drawn in a state where the circuit pattern is precisely arranged with reference to the end faces 39A and 39B of the reticle R.

  Next, an eighth embodiment of the present invention will be described with reference to FIG. FIG. 11 shows a reticle stage RST that supports the reticle R on four mounting portions 10A to 10D, and drive units 12A to 12C that move the reticle stage RST in the X, Y, and θ directions on the reticle base structure CL1. FIG. 5 is a plan view showing the arrangement of a reticle interference system that directly measures the position of the reticle R in the XY direction.

  In the present embodiment, a reticle R having end surfaces 103A and 103B polished by cutting off two corners at an angle of 45 ° is used. Further, a reflection layer is formed on the end surfaces extending in the XY directions that define the remaining two corners 105A and 105B of the reticle R. The interference system according to the present embodiment measures the position change component in the 45 ° direction of each of the two corners 105A and 105B of the reticle R, so that the displacement of the reticle R in the X and Y directions and the rotation in the θ direction. And is configured to measure.

  First, in the first interference system, the laser beam LBa for length measurement is incident on the beam splitter 100A fixed on the reticle base structure CL1 from the 45 ° direction with respect to the XY coordinate system. Of the beam LBa, the beam partially reflected by the beam splitter 100A is perpendicularly incident on the reference mirror 101A fixed on the reticle base structure CL1, and the beam regularly reflected here is the beam splitter 100A and a glass block (cuboid). 102A.

  On the other hand, the beam transmitted through the beam splitter 100A out of the beam LBa is incident on the reticle R in parallel from the oblique end surface 103A of the reticle R, and is formed on two orthogonal end surfaces in the vicinity of the opposite corners 105A. The light is reflected from the back surface by the reflective layer, and is emitted from the end surface 103A at an angle of 45 ° again to reach the total reflection surface 106A of the glass block 102A. The beam reflected by the total reflection surface 106A travels backward along the original optical path, exits from the end surface 103A, is reflected by the beam splitter 100A, passes through the glass block 102A, and is reflected from the reference mirror 101A. The combined beam is an interference beam IBa.

This interference beam IBa is photoelectrically detected by the receiver of the first interference system, and the signal is connected to the counter circuit at the subsequent stage. This first interference system measures the positional change in the 45 ° direction of the corner 105A (vertex) of the reticle R with respect to the total reflection surface 106A of the glass block 102A. The second interference system is configured in exactly the same way, and the length measuring beam LBb is split into two by the beam splitter 100B, and one of the beams is regularly reflected by the reference mirror 101B, and the beam splitter 100B and the glass It passes through the block 102B. The other beam divided by the beam splitter 100B is incident on the reticle R in parallel through the oblique end surface 103B of the reticle R into the reticle R, and is formed on two orthogonal end surfaces near the opposite corner 105B. After being reflected, the light is again emitted from the oblique end surface 103B at 45 ° and reaches the total reflection surface 106B of the glass block 102B.

  The beam reflected by the total reflection surface 106B travels backward along the original optical path and exits from the end surface 103B, is reflected by the beam splitter 100B, passes through the glass block 102B, and is reflected from the reference mirror 101B. It is combined into an interference beam IBb and received by the receiver. This second interference system measures the positional change in the 45 ° direction of the corner 105B of the reticle R with respect to the total reflection surface 106B of the glass block 102B.

  When the first and second interference systems configured as described above are used, the two corners 105A and 105B of the reticle R work as a kind of corner mirror, so that the reticle R can be rotated in the XY plane. There is an advantage that the positions of the reflected beams returning from the total reflection surfaces 106A and 106B of the glass blocks 102A and 102B to the beam splitters 100A and 100B are not shifted.

  In the case of the first and second interference systems described above, the change in position in the 45 ° direction of each of the two corners 105A and 105B of the reticle R is measured, so that each displacement in the X and Y directions is measured. In order to know the quantity, an operation of coordinate transformation is required. However, even when the reticle R is mounted on the reticle stage RST in a relatively large (for example, about 2 ° to 3 °) rotated state, the interference system using the end face of the reticle R functions normally from that state. It becomes possible to make it.

For this reason, the mechanical pre-alignment accuracy before the reticle R is placed on the reticle stage RST is relaxed, and high-speed pre-alignment is possible, so that the effect that the replacement time of the reticle R can be shortened is obtained. It is done.
Next, a ninth embodiment of the present invention will be described with reference to FIG. FIG. 12 is a perspective view showing a configuration of a reticle base structure CL1 and a reticle stage RST suitable for a step-and-scan projection exposure apparatus. In the case of the step-and-scan method, the reticle R is placed on the reticle stage RST as shown in FIG. 12, and the reticle stage RST moves on the base structure CL1 with a stroke corresponding to the dimension of the reticle R (for example, 6 inches). It is configured to move one-dimensionally in the Y direction.

  The movement is performed by linear motor units 200A and 200B arranged extending in the Y direction on both sides of reticle stage RST. Each of the linear motor units 200A and 200B is entirely covered with a cover, and is forced by the exhaust ducts 201A and 201B so that the air in the motor units 200A and 200B heated by the heat generated by the motor coils does not flow outside as much as possible. Exhausted.

  Further, a movable mirror 203 extending in the Y direction is fixed to one end of the reticle stage RST in the X direction, and the movable mirror 203 includes three from the reticle interferometer IFRX for X direction measurement fixed to the base structure CL1. A length measuring beam BMrx is projected vertically. The interferometer IFRX measures a minute position change and a minute rotation error of the reticle stage RST in the X direction (non-scanning direction). Note that a plurality of mesh-shaped openings 200C and 200D are provided immediately below the length measuring beam BMrx of the cover of the linear motor unit 200B, and the air around the length measuring beam BMrx is caused by the exhaust duct 201B. Create a flow that is aspirated to 200D. As a result, the optical path fluctuation of the measuring beam BMrx caused by the cover of the motor unit 200B being warmed is prevented.

  The reticle R is placed on the four placement portions 10A to 10D on the stage RST (the placement portion 10C is a shadow of the reticle R and is omitted in FIG. 12). However, in the present embodiment, in order to prevent the reticle R from shifting in the Y direction (scanning direction) due to acceleration / deceleration of the stage RST during scanning exposure, the end face located at the diagonal portion of the reticle R is loose in the Y direction. A protruding portion is formed on a part of the placement portions 10B and 10D so as to be pressed by the urging force. Further, in the present embodiment, all of the mounting portions 10A to 10D hold the reticle R with a loose vacuum suction force, or only one of the mounting portions 10B and 10D holds the reticle R with a strong vacuum suction force. Configured as follows.

  Further, notch corner portions CM1 and CM2 are formed at two locations on one side extending in the X direction of the reticle R, and are processed at exactly 90 ° within the main surface of the reticle R. The slopes of the corner portions CM1 and CM2 are optically polished, and a reflective layer is deposited on the surface. Therefore, the corner portions CM1 and CM2 act as a kind of corner mirror, and reflect the measurement beams BMry and BMrθ from the Y-direction measuring reticle interferometers IFRY and IFRθ fixed to the base structure CL1. These measurement beams BMry and BMrθ are incident on one of the slopes of the corner portions CM1 and CM2, respectively, and return to the interferometers IFRY and IFRθ in parallel with the incident optical path from the other slope. Thereby, each interferometer IFRY, IFRθ measures the position in the Y direction of each vertex of the two corner portions CM1, CM2.

  As described above, in this embodiment, since the corner portions CM1 and CM2 are formed on one side of the reticle R, the change in the angle of each of the measurement beams BMry and BMrθ caused by the rotation of the reticle R is prevented. There is an advantage that the interference state at each interferometer IFRY, IFRθ is allowed within a certain rotation range of the reticle R. In order to enjoy such advantages, the distances Lk in the X direction of the measurement beams BMry and BMrθ of the interferometers IFRY and IFRθ and the distances in the X direction of the corner portions CM1 and CM2 formed on the reticle R are matched. It is necessary to keep it. In some cases, this means that a dedicated reticle must be prepared for each apparatus manufacturer, but once the standard for the quartz plate for such a reticle is determined, the exposure apparatus manufacturer can easily do that. It is possible.

  Meanwhile, the stage RST shown in FIG. 12 is supported on the base structure CL1 via an air bearing and is moved in a non-contact manner by the linear motor units 200A and 200B. If the position coordinates in the Y direction of the corner portions CM1 and CM2 measured by the two interferometers IFRY and IFRθ are Yc1 and Yc2, the rotational change amount Δθc of the reticle R from the initial state is (Yc1−Yc2). ) / Lk, and the coordinate position Yr of the reticle R in the Y direction at the time of scanning exposure is obtained by calculation of (Yc1 + Yc2) / 2.

  Based on the coordinate position Yr, the linear motor units 200A and 200B are controlled with the same thrust to move the reticle stage RST in the Y direction at a predetermined speed, and the measured rotation change amount Δθc is always a constant value (for example, Zero) or a value that varies with an inherent tendency, and the reticle stage RST may be slightly rotated by giving a difference to the thrusts of the linear motor units 200A and 200B. In FIG. 12, AX is the optical axis of the projection optical system PL, and a rectangular region LE extending in the X direction centered on the optical axis AX is an illumination light irradiation region during scanning exposure.

  According to the above embodiment, since the reticle R is restricted from being displaced in the Y direction by the protruding portions of the mounting portions 10B and 10D that are substantially diagonal, the influence of the expansion of the reticle R is shown in FIG. Appears as a small clockwise rotation of R. However, in this embodiment, since the corner mirror portions CM1 and CM2 are formed at the end of the reticle R, the overall rotation of the reticle R caused by the expansion component of the reticle R and the deviation component on the mounting portions 10A to 10D. It is possible to accurately measure changes over a relatively large range.

  Accordingly, the reticle R and the wafer W are moved relative to each other in the Y direction at a speed ratio equal to the imaging magnification of the projection optical system, and the pattern area portion of the reticle R irradiated with the illumination light LE on the shot area of the wafer W is detected. When the image is scanned and exposed, the relative speed ratio is finely adjusted by an amount corresponding to the amount of expansion of the reticle R from the accurate imaging magnification value, so that the size of the projected pattern region in the scanning direction ( (Magnification) can be adjusted.

  Further, the size adjustment in the non-scanning direction (X direction) of the entire image of the projected pattern area is performed by setting the imaging magnification of the projection optical system PL itself to a very small level according to the expansion amount of the reticle R as described above with reference to FIG. What is necessary is just to respond by changing. Furthermore, in the present embodiment, the projection optical system has an equal magnification, and the mask substrate and the photosensitive substrate are held on a common carriage and the one-dimensional scanning is performed on the equal magnification projection optical system. The present invention can be applied in exactly the same manner to an exposure apparatus that manufactures a large liquid crystal display device, for example, a multi-lens projection exposure apparatus as disclosed in JP-A-7-57986.

  In this embodiment, a temperature sensor for measuring the expansion amount of the reticle R, an interferometer for measuring the reticle dimensions, etc. are not provided. However, if necessary, the expansion amount is indirectly measured as in the previous embodiments. Or you may provide the measurement system which measures directly.

The figure which shows typically the whole structure of the projection exposure apparatus suitable as embodiment of this invention. 1 is a perspective view showing a configuration of a reticle stage and an arrangement of an interferometer system according to a first embodiment of the present invention. FIG. 3 is a plan view showing a configuration of a reticle stage shown in FIG. 2. 1 is a block diagram showing a configuration of a reticle stage control system according to a first embodiment of the present invention. FIG. The perspective view which shows the structure of the reticle stage by the 2nd Embodiment of this invention. The perspective view which shows the structure of the reticle stage by the 3rd Embodiment of this invention. The perspective view which shows the structure of the reticle stage by the 4th Embodiment of this invention. The top view which shows the structure of the interferometer system for expansion amount measurement by the 5th Embodiment of this invention. The figure which shows the typical structure of the projection exposure apparatus by the 6th Embodiment of this invention. The perspective view which shows typically the structure of the sample stage of the EB exposure apparatus with which the reticle manufacturing method by the 7th Embodiment of this invention is implemented. The top view which shows the structure of the interferometer system for reticles by the 8th Embodiment of this invention. The perspective view which shows the structure of the reticle stage RST of the projection exposure apparatus by the 9th Embodiment of this invention.

Explanation of symbols

R ... reticle W ... wafer PL ... projection optical system RST ... reticle stage WST ... wafer stage CL1, CL2, CL3, CL4 ... column structure SX, SY, Sθ ... Reflex part of reticle end face IFRX, IFRY, IFRθ ... Interferometer system for reticle 10, 10A, 10B, 10C, 10D ... Placement part 12, 12A, 12B, 12C ... Reticle drive system

Claims (6)

In an exposure apparatus comprising a stage that is movable while holding a mask substrate, and that exposes a pattern formed on the mask substrate onto a sensitive substrate by a projection optical system,
The exposure apparatus characterized in that the stage holds the mask substrate via an elastic member that can be elastically deformed according to deformation of the mask substrate. The exposure apparatus according to claim 1, further comprising a deformation amount detection device that measures a temperature of the mask substrate and detects a deformation amount due to thermal expansion and contraction of the mask substrate based on the measurement result. The exposure apparatus according to claim 2, further comprising a control device that adjusts an imaging characteristic of the projection optical system based on a detection result of the deformation amount detection device. 4. The exposure apparatus according to claim 3, wherein the control device corrects the position of the mask substrate based on a detection result of the deformation amount detection device. In an exposure apparatus that irradiates a mask substrate on which a pattern is formed with illumination light for exposure, and transfers the pattern onto a sensitive substrate via a projection optical system,
A holding device for holding the mask substrate so as to allow free expansion;
A temperature measuring device for measuring the temperature of the mask substrate;
A detection device for detecting an expansion component based on the measured temperature;
An exposure apparatus comprising: a control device that adjusts an imaging characteristic of the projection optical system based on the calculated expansion component. In a method of forming a circuit pattern by transferring a pattern image in the pattern region onto a sensitive substrate via a projection optical system by irradiating the illumination light for exposure to a pattern region formed on the mask substrate,
Placing the mask substrate on the object plane side of the projection optical system to allow free expansion of the mask substrate;
Measuring the temperature of the mask substrate;
Detecting, based on the measured temperature, an expansion amount from an initial state of the mask substrate that occurs during exposure of the sensitive substrate by the mask substrate;
Imaging characteristics of the imaging projection system in accordance with the detected expansion amount in order to correct minute changes due to expansion of the mask substrate in the state of the pattern image to be exposed in the shot area on the sensitive substrate A circuit pattern forming method comprising the steps of:

Images