JP3384049B2 - Exposure method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an exposure method for transferring an image of a circuit pattern or the like onto a photosensitive substrate, and more particularly to maintaining its image forming characteristics.

[0002]

2. Description of the Related Art Conventionally, as a pattern of a semiconductor integrated circuit is miniaturized, in a projection optical apparatus, a change in imaging characteristics (for example, magnification, focal position) caused by absorption of exposure light by the projection optical system. The need for correction has arisen.
For example, JP-A-60-78455 or JP-A-63
As disclosed in Japanese Patent Laid-Open No. 58349, there is provided a mechanism that detects the amount of light incident on the projection optical system and corrects fluctuations in the optical characteristics of the projection optical system.

To briefly explain this, a model corresponding to the variation characteristic of the image forming characteristic is made in advance, the amount of light energy incident on the projection optical system is obtained by a photoelectric sensor or the like of the stage, and the variation amount is changed with time. Calculate according to this model. That is, the signal of the shutter OPEN during the exposure operation is received, the variation amount of the optical characteristic is constantly calculated according to the model while the exposure is performed, and the correction is performed based on the variation amount.

However, even if the change in the image formation characteristic due to the absorption of the illumination light of the projection optical system is corrected, the mask (reticle) absorbs the illumination light and is thermally deformed, thereby causing the change in the image formation characteristic. There is a problem to say. As a method of correcting the change in the image formation characteristic due to the thermal deformation of the mask as described above, for example, as disclosed in Japanese Patent Laid-Open No. 4-192317, the thermal deformation amount of the mask due to the absorption of illumination light is calculated by a predetermined numerical value. Is calculated, and the change in the imaging characteristic caused according to this thermal deformation amount is calculated. Based on this result, the lens element of the projection optical system is driven in the optical axis direction or in the tilt direction with the axis perpendicular to the optical axis as the rotation axis to correct the variation in the imaging characteristics due to thermal deformation of the mask. A technique for obtaining a desired image formation state is also known.

[0005]

In the conventional technique as described above, it is assumed that the position of the center point of the mask (XY coordinate position) does not move due to the thermal deformation when the thermal deformation amount of the mask is calculated. I am trying. However, in reality, such an ideal thermal deformation rarely occurs due to the influence of the holding member that holds the mask.

The holding member for holding the mask holds the outside of the pattern region (the region where the pattern for actual device manufacturing is formed) from the pattern surface side of the mask by vacuum suction at four points, for example, and holds the projection optical system. Are arranged in a reference plane (XY plane) substantially perpendicular to the optical axis of. Here, the mask shifts in the XY plane direction or rotates in the XY plane due to thermal deformation, for example, when the suction force of the mask of the holding member varies, or when foreign matter exists in the suction portion of the mask. (Rotation). As a result, even if the change in the imaging characteristic due to the thermal deformation of the mask can be corrected, the pattern area of the mask is imaged with a positional deviation or a rotational deviation with respect to the shot area on the wafer. Further, if the position of the mask is sequentially detected and the alignment is performed, the image formation state of the pattern can be always maintained in good condition, but there is a problem that the throughput is significantly reduced.

The object of the present invention has been made in view of the above problems, and corrects the positional deviation of the mask in the reference plane direction due to the thermal deformation of the pattern area so that the image formation state of the pattern of the mask is always good. It is to provide an exposure method that can be maintained at

[0008]

In order to make the present invention easier to understand, the present invention will be described with reference to FIGS. 1 and 5 (a) which is an embodiment. In order to solve such a problem, in the present invention, for example, the pattern area (PA) of the mask (R) held in a predetermined reference plane is irradiated with illumination light (IL) in a predetermined wavelength range. , The mask (R)
In the exposure method of transferring the pattern of FIG. 2 onto the photosensitive substrate (W), thermal deformation of the pattern area (PA) in the mask (R) due to absorption of illumination light (IL) causes the pattern area (PA) to be deformed.
Calculation step of calculating the thermal deformation amount of the pattern area (PA) when the position of a predetermined reference point (RC) in the reference plane in the reference plane is fixed, and a predetermined plurality of points in the pattern area (PA) ( P1, P4, P13, P16),
Range within the reference plane that can move due to thermal deformation (α
1, α2, α3, α4) are respectively determined based on the thermal deformation amount obtained in the calculation step, and a plurality of points (Pa1, Pa4, Pa1) in the pattern region after the thermal deformation obtained in the calculation step are determined.
3, Pa16) are all in the above range (α1, α2, α3,
The process of calculating the maximum value of the displacement amount when the pattern region after thermal deformation obtained in the calculation process so as to exist within α4) is displaced in the reference plane direction, and the mask (R ) Of determining whether to detect the position in the reference plane.

[0009]

[Action] In such present invention, first, the thermal deformation of the mask pattern region for obtaining the reference becomes the thermal deformation amount when it is assumed to have occurred to the position of the predetermined reference point of the pattern area as a fixed, the pattern area You can see the total spread and the shape after thermal deformation. This is calculated based on the assumption that the pattern area does not shift due to thermal deformation, and it is actually unknown at which position in the reference plane the pattern area after thermal deformation is. Therefore, in order to obtain the virtual displacement amount of the pattern area, the range in which a plurality of predetermined points in the pattern area can move by thermal deformation is determined based on the above-mentioned thermal deformation amount, and the heat previously determined within that range is determined. Try to shift the pattern area after transformation. This makes it possible to estimate how much the pattern area after thermal deformation is displaced in the reference plane direction at the maximum. By determining whether or not to perform mask alignment based on this virtual shift amount, it is possible to minimize the number of times of mask alignment.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a schematic configuration of a projection exposure apparatus according to an embodiment of the present invention. In FIG. 1, an illumination light source 1 for exposure such as an ultra-mercury lamp or an excimer laser light source is g
Illumination light IL having a wavelength (exposure wavelength) that sensitizes the resist layer, such as linear rays, i-rays, or ultraviolet pulsed light (eg, KrF excimer laser) is generated. The illumination light IL is
An illumination optical system 6 including an optical integrator (fly-eye lens) and the like after passing through a shutter 2 that closes and opens the optical path of the illumination light, and a semi-transmissive mirror 4 that allows most (90% or more) of the illumination light to pass. Reach Shutter 2
Is driven by the drive unit 3 so as to control transmission and blocking of illumination light. A part of the illumination light reflected by the semi-transmissive mirror 4 enters a photoelectric detector (power monitor) 5 such as a PIN photodiode. The power monitor 5 is the illumination light IL
By photoelectrically detecting the optical information (intensity value) PS to the main control system MCS
Output to. The optical information PS serves as basic data for obtaining the amount of variation in the image forming characteristics of the projection optical system PL in the main control system MCS (details will be described later).

The illumination light IL that has been made uniform in luminous flux and reduced in speckles in the illumination optical system 6 is reflected by the mirror 7, passes through the relay lenses 9a and 9b and the variable blind 10, and then is mirrored. The light is reflected vertically downward at 12 and reaches the main condenser lens 13. Since the surface on which the variable blind 10 is arranged has a conjugate relationship with the pattern area PA of the reticle R, the drive motor 11 opens and closes the operating blades that constitute the variable blind 10 to change the opening position and the shape of the reticle. The illumination field of view of R can be arbitrarily selected. Further, in the present embodiment, the reflected light generated from the wafer W by the irradiation of the illumination light IL passes through the mirror 7 and enters the photodetector (reflection amount monitor) 8. The reflection amount monitor 8 photoelectrically detects the reflected light and outputs optical information (intensity value) RS to the main control system MCS. This optical information RS is the projection optical system P
This is basic data for obtaining the variation amount of the image forming characteristic of L (details will be described later).

The reticle R is held by a reticle holder RH, which is a projection optical system PL.
Is placed on a reticle stage RS that is two-dimensionally movable in a plane (in the XY plane) substantially perpendicular to the optical axis AX.
The coordinate position of the reticle R on the XY coordinate system is sequentially detected by the reticle interferometer 25. The reticle stage controller 27 controls the reticle drive system 26 and the reticle alignment system RA based on the coordinate measurement signal from the reticle interferometer 25 to control the movement and positioning of the reticle stage RS. Initialization of the reticle R is performed by finely moving the reticle stage RS based on a mark detection signal from the reticle alignment system RA that photoelectrically detects the alignment mark AM around the reticle shown in FIG. Then, the reticle R is positioned so that the center point RC of the pattern area PA coincides with the optical axis AX of the projection optical system PL. The reticle R is used after being appropriately replaced by a reticle exchanger (not shown). In particular, when carrying out small-lot production of a wide variety of products, reticles are frequently exchanged.

The illumination light IL that has passed through the pattern area PA enters the projection optical system PL that is telecentric on both sides, and the projection optical system PL projects the projected image of the circuit pattern of the reticle R onto one shot area on the wafer W. The images are superimposed and projected (imaged). The wafer W is mounted on the Z stage 14 which can be finely moved in the optical axis direction (Z direction) by the drive motor 17. Further, the Z stage 14 is mounted on the XY stage 15 which can be two-dimensionally moved in the XY directions perpendicular to the Z direction by the drive motor 18. 2 of XY stage 15
The dimensional position (XY coordinate position) is sequentially detected by the interferometer 19 with a resolution of, for example, about 0.01 μm. Further, a dose monitor 16 is provided on the Z stage 14.
Are provided so as to substantially coincide with the surface position of the wafer W. The light receiving surface of the dose monitor 16 is the projection optical system P.
It has approximately the same area as the L image field (or the projection area of the reticle pattern). This dose monitor 16
Photoelectrically detects the illumination light IL and outputs the light information LS to the main control system MC.
Output to S. The light information LS also serves as basic data for obtaining the amount of variation in the image forming characteristics of the projection optical system PL in the main control system MCS (details will be described later).

Next, the structure of the correction means for correcting the image formation state will be described. In this embodiment, a lens element in the projection optical system PL is driven by a drive element (not shown) capable of expanding and contracting to correct the imaging characteristics (projection magnification, distortion, etc.). Here, regarding this kind of technique, for example, Japanese Patent Laid-Open No.
As disclosed in Japanese Laid-Open Patent Publication No. 21319, a drive element for driving the lens element has a configuration in which the drive element control unit 23 controls the expansion / contraction amount to adjust the optical characteristics of the projection optical system PL. These elements are selected so that the influences on the magnification and distortion characteristics can be controlled more easily than other lens elements. Moreover, it is possible to deal with various shape distortions (trapezoid, rhombus, barrel, pincushion, etc.), and is sufficient for fluctuations in the imaging characteristics of the projection optical system PL caused by thermal deformation of the reticle R due to absorption of exposure light. Can handle. It should be noted that the movement of the lens element is performed within a range in which the influence exerted on other various aberrations (for example, astigmatism) of the projection optical system PL can be ignored. Alternatively, by adjusting the distance between the lens elements, it is possible to adopt a method of correcting the various aberrations while controlling the magnification and distortion characteristics.

The memory 21 shown in FIG. 1 also stores various data necessary for calculating the amount of thermal deformation of the reticle due to absorption of exposure light (type of light blocking member of reticle, pattern density distribution, etc.), and reticle. An equation or a table for calculating the amount of change in the image formation state according to the amount of thermal deformation is stored. And the main control system MCS uses this memory 2
1, power monitor 3, reflection amount monitor 8, irradiation amount monitor 1
The information is obtained from 6, the thermal deformation amount of the reticle R is obtained as described later, and the information is output to the determination unit 28. Further, the main control system MCS obtains a variation amount of the imaging characteristic according to the thermal deformation amount of the reticle R and issues a command to the drive element control unit 23.
Controls the entire device.

Next, a method of calculating the variation amount of the image forming characteristic in this embodiment will be described. In the present embodiment, when calculating the variation amount of the image forming characteristic, first, the thermal deformation amount of the reticle R is obtained. Since it can be considered that the thermal deformation of the reticle R occurs in proportion to the temperature distribution of the reticle R, it is sufficient to know the temperature distribution of the reticle R at a certain time in order to calculate the thermal deformation amount. For example, as a method of simulating this temperature distribution by a computer, there is known a method in which the reticle R is decomposed into certain finite elements and the temperature change at each point is calculated by a difference method, a finite element method, or the like. In this embodiment, the simulation is performed using a relatively simple difference method. First, the pattern area PA on the reticle R is divided into 4 × 4 16 blocks as shown in FIG. 2, and the blocks are designated as blocks B1 to B16. Also, let P be the center point of each block.
1 to P16. The selection of the number of divisions or the calculation method is determined in consideration of the finally required accuracy, the calculation speed of the computer, etc. In the present embodiment, the reticle R is used.
The pattern area PA is simply divided into 16 for convenience.

The reticle R is uniformly illuminated through the illumination optical system 6 when the shutter 2 is open. However, the amount of heat absorbed on the reticle R differs depending on the location depending on the distribution of the pattern of the reticle R. Therefore, the pattern existence rate is obtained for each of the blocks B1 to B16 on the reticle R. At this time, it is assumed that the amount of heat absorbed in each block is uniform.

The pattern existence rate of each block is obtained, for example, by the output ratio between the dose monitor 16 and the power monitor 5 on the Z stage 14. First, the reticle on which no pattern is drawn is positioned so that the center point of the reticle and the optical axis AX of the projection optical system coincide with each other. Next, the XY stage 15 is moved to move the dose monitor 16 to the projection optical system PL.
It is sent to almost the center of. Since the light receiving surface of the dose monitor 16 has substantially the same area as the image field of the projection optical system PL, the dose monitor 16 receives all the exposure light emitted onto the wafer W and photoelectrically detects it. And
The irradiation amount of the exposure light reaching the wafer W via the reticle or the like is calculated. At this time, the irradiation amount monitor 16 is divided into 16 corresponding to the reticle R being divided into 16, and the amount of light passing through each block can be measured independently. At this time, the power monitor 5 also photoelectrically detects the illumination light IL. Next, the same operation as described above is performed on the reticle on which the pattern is drawn. Then, the ratio of the level of the output signal of the dose monitor 16 to the level of the output signal of the power monitor 5 on the reticle on which no pattern is drawn, and the output signal of the dose monitor 16 on the reticle on which the pattern is drawn. The ratio of the pattern on the reticle is calculated based on the ratio between the level of the reticle and the level of the output signal of the power monitor 5.

In the above-mentioned measurement, a reticle having no pattern is used. However, the illumination light is irradiated without the reticle, and the output signal level of the irradiation amount monitor 16 and the output signal level of the power monitor 5 are measured. You may ask for the ratio with. Further, the above-described measurement may be performed every time the reticle is replaced, or may be measured in advance and stored in the memory 21. When the areas of the blocks that divide the reticle R are the same, the size (area) of the light-receiving surface of the dose monitor 16 has a size (area) corresponding to each block, and the pattern existence rate is calculated. At this time, the entire surface of the reticle R may be measured by stepping the stage. Further, if the pattern existence rate is known from the data at the time of manufacturing the reticle, the above measurement is not necessary.

Now, each block B obtained as described above
The heat absorption amount of each block is calculated based on the pattern existence rate of B16. Each block absorbs the amount of heat in proportion to the power of the light source 1 and the pattern existence rate. The absorbed heat escapes in the air or via the reticle holder RH. In addition, heat also moves between the blocks. Here, let us consider the transfer of heat quantity between two objects, for example. It is considered that the movement of the amount of heat in this case is basically proportional to the temperature difference between the two objects. Further, the rate of change of the temperature change due to the movement of the amount of heat is proportional to the amount of movement of the amount of heat. These can be expressed as follows.

ΔQ = K ₁ (T ₁ −T ₂ ) (T ₁ > T ₂ ) (1) dT ₁ / dt = −k ₂ ΔQ (2) dT ₂ / dt = k ₃ ΔQ (3) where ΔQ is the amount of heat transferred, T ₁ and T ₂ are the temperatures of the objects, t is time, and k ₁ , k ₂ and k ₃ are proportional constants.
From the equations (1) to (3), the following equation holds.

DT ₁ / dt = −k ₄ (T ₁ −T ₂ ) (4) dT ₂ / dt = k ₅ (T ₁ −T ₂ ) (5) This is well known. Is a first-order lag system,
When the T _1, T ₂ is a temperature difference, both reach a certain temperature depicts exponential curve. The heat distribution on the reticle R is calculated based on the above equation.

First, focusing on the block B1, this block B1 exchanges heat with the adjacent blocks B5 and B2 (heat transfer). Further, the block B1 also exchanges heat with the surrounding air and the reticle holder 8, but here, in order to simplify the calculation, the temperature of the air and the temperature of the reticle holder 8 are constant. Then, assuming that the temperatures of the blocks are T _{1 to} T ₁₆ , the temperature of the air is T ₀ , and the temperature of the reticle holder 8 is T _H , the following equation holds for the block B1. _{dT 1 / dt = k 12 (} T 2 -T 1) + k 15 (T 5 -T 1) + k 10 (T H -T 1) + k 0 (T 0 -T 1) + k P η 1 P ‥‥ (6 Here, dT ₁ / dt is the time derivative of T ₁ , k ₁₂ and k ₁₅ are coefficients of heat exchange with blocks B1 and B2, blocks B1 and B5, respectively, and k ₀ is each block B1 to B16.
Is the coefficient of heat exchange with the air. Further, η ₁ is the pattern existence rate of the block B1, and P is the power of the light source 1, which corresponds to the output of the power monitor 5. k _P is a coefficient that associates the amount of heat absorbed by each block with illumination light with η ₁ and P. The term of the equation (6) shows the amount of heat absorbed from the illumination light, and the other terms show the process of the absorbed heat being dispersed. Where T _H and T ₀ are constant and T
The temperature of each block can be represented by T ₀ + ΔT with _H = T ₀ , and because each block on the reticle is made of quartz glass, the coefficient of heat exchange between adjacent blocks such as k ₁₂ and k ₁₅ is , Taking into account that they are all equal, equation (6) becomes as follows (where K _R = K ₁₂
= K ₁₃ = ‥ ‥). dΔT / dt = K _R (ΔT ₂ −ΔT ₁ ) + K _R (ΔT ₅ −ΔT ₁ ) + K _H (−ΔT ₁ ) + K ₀ (−ΔT ₁ ) + K _P η ₁ P = (− 2K _R −K _H − K ₀ ) ΔT ₁ + K _R T ₂ + K _R ΔT ₅ + K _P η ₁ P (7) Equation (7) is obtained for each of the blocks B1 to B16, and is expressed in matrix form as follows.

[0024]

[Equation 1]

This is a 16-dimensional simultaneous equation of differential equations of the first order and can be solved by a numerical solution method. Alternatively, the form of differentiation is a minute time (calculation cycle of the calculator)
The value of can be expressed in a differential form and solved.
Since the so-called external force term is the final term in the equation (8), the value of each block per unit time, that is, η ₁ P ₁ , η ₂ P ₂
If we put the value of ... into the calculator, ΔT _{1 for} each time,
The value of ΔT ₂ can be calculated. The pattern existence rates η ₁ , η _2, ... Are obtained by actual measurement as described above,
The incident light amounts P ₁ , P _2, ... Are obtained by the power monitor 5 and the irradiation amount monitor 16.

The coefficients K _R , K ₀ , K _H and K _P can be calculated from the reticle, the physical properties of air, the flow velocity of air and the like. Alternatively, it is possible to perform experiments on various reticles and determine each coefficient so that it actually best fits. From the above, the temperature distributions ΔT _{1 to} ΔT ₁₆
Is required. From these and the coefficient of expansion of the quartz glass, the mutual distance change between the center points P _{1 to} P ₁₆ of the blocks B _{1 to} B ₁₆ can be obtained, and the movement of each point on the reticle can be determined. Based on this, it is possible to calculate the fluctuation of the imaging characteristic, for example, the distortion of the image projected on the wafer W.

The method up to the above is such that once the temperature distribution of the reticle is
After obtaining ΔT, the movement of the center point P is obtained, and the image distortion is obtained.
However, the direct image distortion (data
Distortion, field curvature, etc.)
is there. At this time, each coefficient K _R, K₀, K_H, K_PThe real
Includes changes in deflection of reticle R if obtained by test
It will be shaped like Moreover, the thermal conductivity of the reticle R is very good.
If there is only a part of the pattern, or only part
Consider that the reticle R expands even when exposed to light.
However, if there is no problem in accuracy, the complicated
No calculation is required, and simpler calculation is sufficient.

Next, the operation of this embodiment will be described. First, it is necessary to determine each parameter of the equation (8) for calculating the fluctuation of the image formation state due to the thermal deformation of the reticle R at the time of manufacturing the device and store it in the memory 21. The parameter K _R relating to heat conduction of the reticle _R is an amount determined by the material, thickness, etc. of the glass which is the main material of the reticle R.
This can be determined by physical property values or experiments as described above. Further, K ₀ and K _H are parameters relating to heat conduction between the glass substrate and the reticle holder RH or air, and these are also quantities determined by the material of the glass substrate. Next, K _P , which is a parameter relating to heat absorption of the light-shielding portion of the reticle R and the glass material, is determined by the materials of the light-shielding member of the reticle and the glass substrate.

These parameters need to be stored in the memory 21 in the form of a function of physical properties of the material of the reticle R, such as glass, or in the form of a table or the like, but they have an effect on the final image formation state. It may be a constant value as long as is negligible. When the parameters are obtained by an experiment, the projection optical system PL itself also absorbs the illumination light. Therefore, the thermal deformation of only the reticle R is measured as follows. First, when both the reticle R and the projection optical system PL are sufficiently in equilibrium with the outside temperature, the image of the reticle R is exposed on the wafer W, and then a light shield is inserted between the reticle R and the projection optical system PL. The shutter 2 is opened for a certain period of time to irradiate the illumination light. Next, the shading object is removed, the image of the reticle R is exposed again on the wafer W, and the amount of change in the image formation state due to thermal deformation of the reticle R can be known by comparing with the image exposed first. When the thermal deformation of the reticle R has been sufficiently corrected by this method, the illumination light absorption of the projection optical system PL can be detected by irradiating the reticle R without a light shield. As a result, the thermal change between the reticle and the projection optical system can be separated and corrected, and even if the reticle is replaced during operation, the correction can be accurately performed. The absorption of the illumination light of the projection optical system PL can be corrected by the conventional technique.

As described above, in order to calculate the amount of thermal deformation of the reticle R, the glass material of the reticle R, the type of light shielding material, the pattern existence rate of each block, and the numerical aperture of the variable blind 10 are required. Of these, the attributes of the reticle R are measured for each reticle, or are stored in advance in the memory 21.
Stored in. The heat absorption coefficient of chromium, which is the light shielding portion, can be estimated to some extent from the reflectance of chromium. First, a reflecting surface having a known reflectance is provided on the stage side (lower side in FIG. 1) of the projection optical system, and the output signal of the reflection amount monitor 8 at that time is stored. The reflectance can be calculated by obtaining the reflected light component from the reticle surface from the output signal, the reticle pattern existence rate obtained in advance, and the reflectance components of other lens members and the like.

During the exposure operation, the light information (intensity values) LS, RS and PS from the power monitor 5, the reflection amount monitor 8 and the irradiation amount monitor 16 are output to the main control system MCS, respectively. The main control system MCS calculates the change in the image formation state due to the heat absorption between the reticle R and the projection optical system PL from these information and each data stored in the memory 21, and the total change amount is calculated. calculate. When the imaging state of the projection optical system PL changes due to other factors such as atmospheric pressure change, these changes are also totaled.

FIG. 3 shows a lattice (hereinafter referred to as an ideal lattice) formed by the center points (P1 to P16) of each block of the pattern area PA of the reticle R, and after the pattern area PA obtained by the above calculation is thermally deformed. Grids (hereinafter referred to as reticle grids) consisting of the center points of the respective blocks Pa1 to Pa1
It is the figure which represented 6. A region α1 on the XY coordinate system in FIG. 3 is a region on the opposite side to the side where the ideal lattices (P1 to P16) exist with the X coordinate value of the point P1 as a boundary in the X direction, and in the Y direction. Is an area on the opposite side of the side where the ideal lattice exists with the Y coordinate value of the point P1 as a boundary. Similarly, with respect to each of the areas α2, α3, and α4, similarly, with respect to the X direction, it is an area on the side opposite to the side where the ideal lattice exists with the X coordinate value of the points P4, P13, and P16 as a boundary, and also in the Y direction. This is a region opposite to the side where the ideal lattice exists with the Y coordinate value of each point as a boundary. As shown in FIG. 3, the main control system MCS
Is a thermal deformation amount on the assumption that the thermal deformation of the reticle R occurs with the coordinate position of the center point RC of the pattern region fixed. This serves as a reference when correcting the variation amount of the image formation characteristic due to the thermal deformation of the pattern area by the lens element in the projection optical system PL. Therefore, the amount of thermal deformation is obtained with reference to the point where the optical axis AX of the projection optical system PL and the reticle R intersect, that is, the center point RC of the pattern area.

Next, the main control system MCS calculates an optimum correction amount for the image formation state of the pattern of the reticle R after thermal deformation. Then, a command is output to the drive element control unit 23 to drive the lens element in the projection optical system PL to correct the image formation state. The correction method is described above in Japanese Patent Laid-Open No. 5-2131.
Since it is described in detail in Japanese Patent Publication No. 9, the description is omitted here.

Now, the main control system MCS corrects the image forming characteristics of the projection optical system PL, and outputs information on the amount of thermal deformation of the reticle to the judging section 28 as shown in FIG. The determination unit 28 is a range in which the vertices P1, P4, P13, P16 of the ideal lattices P1 to P16 of the pattern area PA can be moved by thermal deformation based on the thermal deformation amount (spreading amount) of the ideal lattice obtained by the above-described calculation. (Area) α1, α2, α
3 and α4 are determined. This is because the thermal deformation obtained by the calculation is deformed in the direction in which it extends in the entire pattern area, so that even if the reticle gratings Pa1 to Pa16 are displaced due to the thermal deformation, the vertices P1, P4, P1
3, P16 are set based on the premise that they are not deformed (moved) in the contracting direction with respect to the X direction or the Y direction. However, in the present embodiment, for the sake of convenience, the region is defined as shown in FIG. 3, and the region is not limited to this region. The determining unit 28 then determines the four vertices Pa1, Pa4,
Pa13 and Pa16 are all areas α1, α2, and α, respectively.
3, reticle gratings Pa1 to Pa so as to exist within α4
When 16 is shifted in the XY plane direction, a value that maximizes the amount of deviation is calculated. The state of the reticle grating at that time is shown in FIG.

FIG. 4A shows the reticle grating (Pa1 to Pa16) shown in FIG. 3 and this reticle grating (Pa1 to Pa16).
Pa16) is a diagram showing a state when it is moved in the XY directions, a point Ps1 exists at the apex of the area α1 (the point where the ideal lattice P1 existed), and Ps13 is on the boundary line in the Y direction of the area α3. Exists. Further, Ps4 and Ps16 exist within the range of the areas α2 and α4. In addition, assuming that the thermal deformation of the reticle R occurs at a fixed coordinate position of a point other than the center point RC of the pattern area, the reticle lattice (P
The shape (size) of s1 to Ps16) is the same as the shape of the reticle grating (Pa1 to Pa16) when it is thermally deformed with the coordinate position of the center point RC of the pattern region fixed.
Then, the determination unit 28 determines the X of these two reticle gratings.
The relative displacement amount (shift amount) Ds in the Y plane direction is obtained.

Further, in this embodiment, the rotation due to the thermal deformation of the reticle is also taken into consideration. FIG. 4B shows a reticle lattice (points Pr1 to P1) that maximizes the rotation amount, assuming that the reticle after the deformation is rotated about the center point RC of the pattern area PA in the XY plane.
r16) is shown. Here, the point Pr16 is on the boundary line extending in the X direction of the region α4. Similar to the above, the determination unit 28 determines the four vertices (Pa1, Pa4, Pa13, Pa16) of the reticle lattice after thermal deformation obtained by the calculation and the four vertices of the reticle lattice when the reticle is rotated ( Pr1, Pr4, Pr13, Pr1
A deviation amount Dr in the XY directions from 6) is obtained.

These deviation amounts Ds and Dr are the four vertexes Pa1, Pa4, Pa13 and Pa1 of the reticle after deformation.
6, and the coordinate positions of the center point RC and the conditions of the regions α1 to α4, which are the movable ranges of the respective vertices on the stationary coordinate system, can be easily obtained. By the way, the allowable range of the reticle (reticle grating) deviation amount that does not affect the image formation characteristic of the pattern is stored in the determination unit 28 in advance. Then, when at least one of the calculated deviation amounts Ds and Dr exceeds the allowable range, the determination unit 28 outputs a command signal to the reticle stage controller 27 to perform reticle alignment. In addition to the command signal, the determination unit 28 also outputs, to the reticle stage controller 27, information regarding the shape (size) of the reticle grating after thermal deformation received from the main control system MCS.

When the reticle stage controller 27 receives the command signal, it controls the reticle alignment system RA to detect the reticle mark AM provided on the reticle R. The shape (size) of the reticle after thermal deformation is the determination unit 2
The reticle stage controller 27 and the reticle alignment system RA perform the alignment of the reticle R based on the shape of the reticle after the deformation, as can be seen from the information about the reticle grating received from the reticle. Here, the actual displacement amount of the reticle R must be within the above-mentioned allowable range, or the position of the reticle R must be near the initial setting position (the position where the center point of the reticle R and the optical axis AX of the projection optical system coincide). For example, it is not necessary to move the reticle stage RS to align the reticle R with the initial setting position.

In this way, the reticle R that absorbs the illumination light is used.
Is corrected by moving the lens element in the projection optical system PL, and the reticle shift or rotation due to the thermal deformation is corrected by reticle alignment. Thus, the pattern areas of the reticle after thermal deformation can be more accurately overlapped. Further, when the reticle alignment is performed, the main control system MCS moves the movable ranges α1, α2, α3, α of the vertices of the reticle lattice.
4 is newly set. At this time as well, since the position of the reticle after alignment is known, the region is defined in the same manner as the above-mentioned conditions with each vertex of the reticle grid as a reference. In addition, when the irradiation of the reticle with the illumination light is stopped for a long time or the intensity of the illumination light is reduced, the thermal deformation amount of the reticle lattice obtained by the calculation may be negative (deformation that shrinks). . In this case, the movable range α1, α2, α3, α4 of the four vertices of the reticle lattice is symmetrical with respect to the vertices P1, P2, P3, P4 of the ideal lattice (the ideal lattice exists in both the X and Y directions). The reticle deviation amount can be estimated by the same method as the above-described embodiment.

In this embodiment, the above calculation is performed at regular time intervals from the start of exposure until the reticle is thermally stabilized to update the data relating to the thermal deformation amount of the reticle and the reticle displacement amount. The reticle may be aligned by driving the lens element group based on the updated data. In addition, the timing for performing the above calculation may be arbitrary, and may be appropriately determined depending on, for example, the opening / closing time of the shutter 2. Alternatively, the above calculation may be always performed from the start of exposure.

If it is assumed in FIG. 4 (b) that the deformed reticle is rotated, the determination unit 28 calculates the rotation amount (rotation angle) θ instead of calculating the deviation amount Dr. Good. At this time, the determination unit 28 determines the allowable range for the shift amount (positional shift amount) when the reticle is shifted, and the rotation amount (rotational shift amount) when the reticle is rotated.
The permissible range for each is stored individually and compared with each other. Further, in this embodiment, the shift amount Ds of the reticle shown in FIG. 4A may be simply obtained without considering the rotation of the reticle.

Further, in the present embodiment, the above-mentioned shift amount is calculated separately when the reticle only shifts and when it only rotates, but both shift and rotation occur. It is also possible to obtain the amount of reticle displacement when it is assumed that The deviation amount of the reticle here means each vertex (Pa1, Pa4, Pa) of the reticle lattice after thermal deformation obtained by calculation.
13, Pa16), and the vertices (Pa1 ′, Pa4 ′, Pa13 ′, Pa1) of the reticle lattice after thermal deformation assuming that both shift and rotation occur.
6 ') is the maximum value among the deviation amounts.

In this embodiment, the reticle R is set to 1
Although it is divided into six parts and the thermal deformation amount at the center point of each block is obtained, it can be said that the integrated irradiation energy of the illumination light with which the reticle is irradiated and the thermal deformation amount of the reticle are in a substantially proportional relationship. A certain amount of thermal deformation of the reticle can be obtained from the integrated irradiation energy of the irradiation light. Then, the correspondence relation between the thermal deformation amount of the reticle and the displacement amount due to the thermal deformation of the reticle, and the allowable range of this displacement amount are stored in the memory 21 in advance,
It is possible to sequentially measure the integrated irradiation energy applied to the reticle during the exposure operation and determine whether to perform reticle alignment based on the value of the integrated irradiation energy. As described above, it is possible to easily estimate the shift amount when it is assumed that the mask shifts in the reference plane direction due to the thermal deformation of the mask due to the absorption of the illumination light.
Since the position of the mask in the reference plane can be detected only when the amount of this deviation has a value that affects the image formation of the pattern on the photosensitive substrate, the number of times of detecting the position of the mask can be minimized. It is possible to limit it.
Then, when the actual mask position is deviated to such an extent that it adversely affects the image formation of the pattern, the deviation is corrected, so that the overlay accuracy of the pattern images is further improved without significantly reducing the throughput. Higher precision imaging can be performed.

[0044]

As described above, according to the present invention, the mask is formed only when the amount of the mask displacement due to the thermal deformation due to the absorption of the illumination light of the mask has a value that affects the image formation of the pattern on the photosensitive substrate. Since it is only necessary to detect the position of, the number of times of detecting the position of the mask can be minimized. Further, when the mask position is displaced to such an extent that it adversely affects the image formation of the pattern, the displacement is corrected,
The overlay accuracy of the pattern images is further improved without significantly reducing the throughput, and higher-precision imaging can be performed.

[Brief description of drawings]

FIG. 1 is a diagram showing a schematic configuration of a projection exposure apparatus according to an embodiment of the present invention.

FIG. 2 is a top view showing a state in which a pattern area PA on a reticle is divided into 16 areas.

FIG. 3 is a diagram schematically showing thermal deformation of a reticle.

FIG. 4A shows a reticle grating when the amount of shift of the deformed reticle in the X and Y directions is maximum, and FIG.
FIG. 6B is a diagram showing a reticle grating when the rotation amount of the deformed reticle in the XY plane is maximum.

[Explanation of symbols]

1 ... Light source 21 ... Memory 27 ... Reticle stage controller 28 ... Judgment unit W: Wafer R ... Reticle PL: Projection optical system RA: Reticle alignment system MCS ... Main control system

Continuation of front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) H01L 21/027 G03F 9/00

Claims (10)

(57) [Claims] 1. An exposure method for irradiating a pattern area of a mask held in a predetermined reference plane with illumination light in a predetermined wavelength range to transfer the pattern of the mask onto a photosensitive substrate. The thermal deformation of the pattern area due to absorption of
A calculation step of calculating a thermal deformation amount of the pattern area when the position of the predetermined reference point in the pattern area in the reference plane is fixed, and the heat of the plurality of predetermined points in the pattern area. The range in the reference plane that can be moved with the deformation is determined based on the thermal deformation amount obtained in the calculation step, and the plurality of points in the pattern area after the thermal deformation obtained in the calculation step are respectively defined. A step of calculating a maximum value of the deviation amount when the pattern area after thermal deformation obtained in the calculation step so as to be present in the range is shifted in the reference plane direction, and based on the value of the deviation amount And a step of determining whether to detect the position of the mask in the reference plane. 2. An exposure method for exposing a substrate by projecting an image of a pattern of the mask onto the substrate via a projection optical system, in order to obtain the amount of displacement of the mask, a mask on the mask is obtained.
An exposure method, characterized in that whether or not to detect a mark is determined based on a thermal deformation amount of the mask. 3. The method according to claim 2, wherein an amount of misalignment of the mask caused by thermal deformation of the mask is estimated, and whether or not to perform alignment of the mask is determined according to the amount of misalignment. The method described. 4. An exposure method for exposing a substrate by projecting an image of a mask pattern onto the substrate via a projection optical system, wherein an amount of displacement of the mask due to thermal deformation of the mask is estimated. , Of the mask according to the estimated displacement amount
Detection of mark on the mask in order to determine the actual amount of deviation
An exposure method, characterized in that it is determined whether or not light is emitted . 5. The amount of displacement of the mask is estimated based on the intensity of illumination light with which the mask is projected to project an image of the pattern of the mask onto the substrate. The method according to 3 or 4. 6. The method according to claim 3, wherein the shift amount of the mask is estimated based on the intensity of reflected light generated from the substrate. 7. An exposure method for exposing a substrate by projecting an image of a pattern of the mask onto a substrate via a projection optical system, wherein an amount of displacement of the mask due to thermal deformation of the mask is predetermined. When it is estimated that the allowable range is exceeded, the mass
Mark to detect the actual amount of misalignment of the mask.
Exposure method which is characterized in that out. 8. The actual displacement of the mask is
If it exceeds the allowable range, correct the mask deviation.
The method according to claim 7, wherein: 9. The mask is displaced without rotating the mask.
It contains this, The any one of Claims 2-8 characterized by the above-mentioned.
The method described in. 10. The putter that accompanies the thermal deformation of the mask
Changes in the image formation state of the projection image of the projection optical system.
Claim that is corrected by adjusting
Item 10. The method according to any one of Items 2 to 9.