JPH10135119A - Projection aligner - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a semiconductor aligner, which can absorb illumination light by a reticle and can quickly correct a deformation too asymmetrical to be rearranged by a projection lens by removing the deformation. SOLUTION: A projection aligner illuminates a circuit pattern on an original plate which illumination light to project the pattern light on a substrate to be exposed through a projection lens. The aligner includes four or more position measuring marks provided on the original plate for the purpose of detecting a thermal deformation of the original plate caused by absorbing the illumination light, four or more reference marks provided on a base disk carrying the original plate thereon, as opposed to the position measuring marks, and means for measuring the shifts in the coordinate values of the position measuring marks with respect to the reference marks, and for correcting a change in the pattern projected on the substrate on the basis of the measured shifts.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for projecting an electronic circuit pattern on a reticle surface to each shot on a wafer surface via a projection optical system in order when manufacturing semiconductor devices such as ICs and LSIs. In particular, in the case of an exposure apparatus (a so-called stepper), even when a reticle absorbs exposure light and thermal expansion occurs during projection exposure, thermal stress generated in the reticle is always minimized, and a magnification component generated by expansion is immediately reduced. More specifically, the present invention relates to an exposure apparatus for semiconductor manufacturing having a function of correcting an error.

[0002]

2. Description of the Related Art In recent years, as patterns of semiconductor integrated circuits such as ICs and LSIs have become finer, projection exposure apparatuses are required to further improve resolution, overlay accuracy, and throughput. Especially for overlay accuracy,
In recent years, the problem that the reticle absorbs the exposure light to cause a chip magnification and a focal point shift has become apparent. Since the reticle uses quartz glass, the absorptivity of the exposure light (KrF 243 nm, i-line, g-line, etc.) of the glass itself is about several ‰, and its linear expansion coefficient is 0.5 pp.
Because of its low m / ° C., the thermal expansion of the conventional reticle has hardly been a problem. However, the existence ratio of Cr patterns (area ratio) has been increased due to demands for higher throughput, such as higher luminance of illumination light lamps and three layers of Cr surfaces for preventing flare of optical systems. When the value is high, the exposure energy absorbed by the pattern also increases, and the phenomenon that the temperature of the quartz glass rises due to the heat conduction process and thermal expansion occurs has been confirmed. That is, when the exposure is started, the reticle Cr surface generates heat by absorbing the illumination light, and the internal heat is conducted into the quartz glass of the reticle. As a countermeasure, there is a method of preventing temperature rise by blowing air conditioned to the reticle.However, it is impossible to make the temperature of the air and the temperature of the reticle the same, the device becomes large, This is not very realistic because the effect is low with a reticle equipped with a pellicle. Therefore, as disclosed in Japanese Patent Application Laid-Open No. 4-192317, a method for correcting thermal deformation of a reticle using a projection optical system has been considered.

[0003] This method will be briefly described. The amount of thermal deformation of a reticle due to absorption of illumination light is determined by the heat absorption rate of the reticle,
Alternatively, it is obtained by numerical calculation based on the density distribution of the pattern or by directly measuring the deformation of the reticle. Based on this, the amount of change in the imaging state of the projection optical system is predicted by calculation. Based on the result, the correction is performed using the image forming state correction means of the projection optical system.

[0004]

However, the above prior art is insufficient in the following three points.

First, when the reticle Cr pattern abundance is uneven, the amount of heat generated varies depending on the location, and the amount of deformation also varies depending on the location. Alternatively, if the vacuum pressure when the reticle is fixed to the reticle stage and the contact state thereof differ at each vacuum point, the reticle may not expand at the center of the reticle. Therefore, if the projection lens is corrected while the reticle is in such a state, an overcorrected area and an insufficiently corrected area are formed in the angle of view, and in some cases, the correction is not performed. However, the accuracy of the superposition may be improved.

Second, when the reticle thermally expands, there is a possibility that the reticle may expand as if the bowl were lying down, depending on the state of the vacuum. In our simulations, it was found that when the reticle was completely locked to the reticle stage, the bowl-like deformation occurred. The expansion and contraction of the reticle pattern in the x and y plane directions can be corrected by a known magnification correcting means of the projection lens. However, there is currently no effective correction means for the bulge in the z direction.

Third, considering that the stepper is required to have a higher throughput, the amount of thermal deformation of the reticle is calculated or directly measured, and the correction amount of the magnification correction function is calculated based on the calculated amount. Therefore, if the projection lens is corrected, it is disadvantageous for this requirement.

SUMMARY OF THE INVENTION In view of the above, it is an object of the present invention to eliminate uncorrectable deformation even if a reticle absorbs illumination light and asymmetric deformation occurs, thereby enabling a quick correction operation by a projection lens. It is an object of the present invention to provide a semiconductor exposure apparatus for the purpose.

[0009]

In order to solve the above-mentioned problems, the present invention illuminates a circuit pattern on an original plate with illumination light having a predetermined wavelength, and performs projection exposure on a substrate to be exposed via a projection lens. In the projection exposure apparatus, in order to detect the thermal deformation generated by the original plate absorbing the illumination light,
Four or more position measurement marks are provided on the original plate, and at the same time, four or more reference marks are provided on the base board supporting the original plate so as to face the position measurement marks. It is characterized in that a means is provided for measuring a deviation of the coordinate values and correcting a change in a pattern projected on the substrate to be exposed in consideration of the deviation. The means for correcting this change in the pattern usually consists of the following two sequences.

First, when the amount of thermal deformation of the original measured by the measuring means exceeds a predetermined allowable value, that is, when it is determined that the level has become a problem in the actual exposure process, the projection exposure is performed. The operation is stopped once, the vacuum holding the original plate on the stage that supports the original plate and can move parallel to the base plate is released, the thermal stress is released, and the original plate is chucked again, and each position is fixed. The stage is aligned in the X, Y, and θ directions so as to allocate the marks so that the sum of squares of the thermal deformation errors generated in the measurement marks is minimized. Thus, once the vacuum is released, the thermal deformation caused by the influence of the illumination light is reduced to X and Y.
It can be limited only to the direction.

Second, a magnification component generated on the original plate,
The correction amount of the magnification correction function of the projection lens corresponding to the magnification component is stored in a storage device in advance as a table, and from the coordinate values of each position measurement mark, the magnification component due to thermal deformation of the original plate is converted. The magnification correction is performed by obtaining the correction amount of the corresponding magnification correction function from the table. Here, the magnification component held in the table (reticle thermal deformation amount-projection lens magnification correction amount table) and the corresponding correction amount of the magnification correction function are obtained in advance by experiments, simulations, and the like. Further, after the magnification correction of the original plate is performed, the coordinate value of each position measurement mark may be used as the initial coordinate of the reticle in the restarted exposure operation.

After the above operation is completed, the exposure operation is restarted again. Thereafter, the reticle position measuring means constantly monitors the reticle deformation amount, and when it is determined that the reticle deformation amount has reached a problematic level, the same operation is repeated again.

With the above arrangement, even if the reticle absorbs illumination light and asymmetrical deformation occurs, it is possible to eliminate uncorrectable deformation as needed, thereby enabling quick magnification correction operation by the projection lens. And a good circuit pattern can always be printed on the wafer.

[0014]

BEST MODE FOR CARRYING OUT THE INVENTION

(Embodiment 1) Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

First, before describing the main part of the present invention, the configuration of a projection exposure apparatus (stepper) to which the present invention is applied will be briefly described with reference to FIG.

In this figure, reference numeral 60 denotes an illumination system, 1 denotes a reticle on which a circuit pattern is formed, and 6 denotes a reticle stage substrate on which a reticle stage (not shown) is mounted. Reticle 1 is positioned with reference to reticle stage substrate 6. The measured value of the positioning is stored in the console unit 71 which controls the entire apparatus.

Reference numeral 61 denotes a pattern image of the reticle 1 on the wafer 6.
Reference numeral 4 denotes a projection lens for projecting, and 62 denotes a lens drive unit for correcting a change in the imaging performance of the projection lens 61 due to a known atmospheric pressure and temperature. 72 is off-axis and T
A mirror for bending a probe light for wafer alignment measurement by a TL (Through The Lens) method, and 69 is a measurement system thereof.

Reference numerals 63 and 70 denote well-known focus and wafer tilt detectors, which irradiate a light beam on the surface of the wafer 64 (63) and photoelectrically detect the reflected light (7).
0), the in-focus position of the projection lens 61 and the inclination of the wafer 64 are detected.

Reference numeral 65 denotes a wafer chuck for vacuum chucking the wafer 64, and 66 denotes a wafer stage capable of coarse and fine movement in the x, y, and θ directions, the position of which is constantly monitored by an interferometer mirror 67 and an interferometer 68. I have. The console unit 71 controls the entire exposure apparatus including these main units.

FIG. 3 shows a reticle according to the present invention, 1 is a reticle whose circuit pattern is not shown, and 2 to 5 are marks for measuring the position of the reticle, respectively. FIG. 4 is a diagram showing reference marks provided at positions facing the four marks. Reference marks 2a to 5a and reticle position measurement marks 2 to 5 are paired. Reference numeral 6 denotes a substrate that supports a reticle stage (not shown). FIG. 5 shows the positional relationship between the reticle 1, the reticle stage substrate 6, the reticle position measurement marks 2 to 5, and the reference marks 2a to 5a.
Shown in However, the reticle stage is not shown in FIG. As can be seen from FIGS. 2 to 5, the reticle position is obtained by measuring the reticle position measurement marks 2 to 5 with reference to the reference marks 2a to 5a by measuring means (not shown). The positions of the reticle position measurement marks 2 to 5 can be detected as x and y coordinates, and each measurement value is (x ⁽ⁱ⁾ _j , y ^{(i ) )} _j ).

Prior to exposure, the reticle 1 having the four reticle position measurement marks as described above is chucked and aligned on the reticle stage with reference to the reference marks on the reticle stage substrate 6. When the exposure is started in such a state and the exposure is performed intermittently, the Cr pattern portion of the reticle absorbs the exposure light, so that the temperature rises and thermal deformation occurs. As described above, quartz glass, which is the base material of the reticle, also has a linear expansion coefficient of 0.1.
As low as 5 ppm, Cr pattern abundance (area ratio)
When is low, thermal expansion of the reticle is of little concern. However, when the abundance ratio (area ratio) of the Cr pattern is high, the exposure energy absorbed by the pattern increases, and the temperature of the quartz glass also increases due to the heat conduction process, and the quartz glass expands thermally. When the temperature of the reticle is plotted with the horizontal axis representing time, the results are as shown in FIGS.
FIG. 12 shows a case where the existence ratio of the Cr pattern is high, and FIG. 13 shows a case where it is low. The temperature rise of the reticle also follows the normal transient heat transfer process. That is, when the exposure is started at τ = 0, the temperature gradually increases, and the error function e
Follow the curve of rf (ε). At a certain time τ = τ0, the steady state of the maximum temperature の = Τ0 is reached. When the exposure is stopped at τ = τ1, the curve follows the reverse curve to that at the time of the temperature rise, cools down, and again reaches the steady state. The key to this process is
This is a section (0 ≦ τ ≦ τ0) of an unsteady state from the start of exposure to reaching a steady state. In the present embodiment, the main focus is on appropriately detecting the thermal expansion of the reticle in this section and performing efficient correction.

FIGS. 6 and 7 show the reticle in a state where a certain time has passed since the start of the exposure. When the existence ratio of the circuit pattern is substantially uniform on the reticle surface, the temperature distribution due to the absorption of the exposure light becomes substantially concentric as shown in FIG. 6, and the resulting thermal deformation becomes vertically and horizontally symmetrically. In the figure, the dotted line represents the reticle before deformation, 11 is the deformed reticle, and 13 to 16 are the moved reticle position measurement marks. Further, when the existence ratio of the circuit pattern is deviated on the reticle surface, it is easily expected that the temperature distribution is deviated as shown in FIG. 7 and the accompanying thermal expansion does not involve symmetry. In FIG. 1, reference numeral 11a denotes an asymmetrically deformed reticle, and reference numerals 13a to 16a denote asymmetrically moved reticle position measurement marks. The means of the present embodiment for effectively correcting the deformed reticle will be described below.

In this embodiment, these deformation states are monitored by periodically measuring the reticle position measurement marks. Maximum deviation of the reticle position measurement mark from the initial coordinate value max. Δx ⁽ⁱ⁾ _j or max. Δy ⁽ⁱ⁾ _j is
Exceeding process line width to be symmetrical, superposed preset reticle thermal deformation amount is the allowable value e _x or e _y from accuracy, temporarily stops the exposure. Then, after the reticle vacuum is released and the thermal stress is released, chucking is performed again. This vacuum release process eliminates inadvertent deformation of the reticle in the z direction and reduces the deformation to x,
Necessary to limit in y-direction. Since the reticle is chucked again in a thermally expanded state, it is necessary to perform reticle alignment again. 9 and 10 show the alignment method.

In FIG. 9, reference numerals 2a to 5a are reference marks as described above, and reference numerals 13a to 16a are reticle position measurement marks which are out of position. Then, the displacement of the reticle position measurement mark from the reference mark is calculated by (Δ
x ⁽ⁱ⁾ _{j + 1} , Δy ⁽ⁱ⁾ _{j + 1} ). In the reticle alignment of the present invention, the measurement marks are allocated so that the positional deviation from each reference mark is minimized. That is, as shown in FIG. 6, when thermal deformation occurs as a concentric circle around the center of the reticle, the amount of misalignment allocated to each position measurement mark is the same, but when the reticle deforms asymmetrically as shown in FIG. Since the amount of displacement differs depending on the position of each position measurement mark, the sum of squares of the displacement is

[0025]

(Equation 1) The reticle stage is driven by x, y, and θ so that is minimized, and alignment is performed. FIG. 10 shows the arrangement of the measurement marks as a result of the alignment. As can be seen from a comparison with FIG. 9, the marks 14a and 15a having large displacements have smaller displacements like 14b and 15b, and the marks 13a and 16a having smaller displacements have 13b and 15b.
16b, the displacement is large. But,
The maximum value of the displacement max. Δx ⁽ⁱ⁾ _j and max. Δ
y ⁽ⁱ⁾ _j can be reliably reduced. This is effective when the magnification correction means 62 of the projection lens 61 corrects a rotationally symmetrical magnification component with respect to the lens optical axis, but the non-rotationally symmetrical magnification component is difficult to correct. Is based on That is, correction is performed after approximating more rotationally symmetric deformation by allocating the non-rotationally symmetric thermal deformation error generated in the reticle to each mark. FIG. 8 shows the reticle thus aligned.

Next, a method of correcting the thermally expanded reticle using the magnification correcting means 62 of the projection lens 61 will be described. When the reticle alignment is completed, the reticle has a magnification component corresponding to the deviation (Δx ⁽ⁱ⁾ _{j + 2} , Δy ⁽ⁱ⁾ _{j + 2} ) from the initial coordinates of each reticle position measurement mark. Become. In the present embodiment, the displacement amount is integrally expressed as a magnification in ppm, and a correction operation of the projection lens is performed based on the magnification component. The correction amount is obtained in advance by experiments and simulations, and a table defining the correction amount of the magnification correction unit 62 for the magnification generated on the reticle is prepared on the console 71. In the actual sequence, the magnification component generated by the reticle 1 selects a correction amount corresponding to the magnification component from the table on the console 71, the command value is input to the magnification correction unit 62, and the lens is driven and corrected.
Is controlled by the open loop. Once the reticle magnification is corrected in this way, the deviation of each reticle position measurement mark from the initial coordinates at the present time (Δx ⁽ⁱ⁾ _{j + 2} ,
Δy ⁽ⁱ⁾ _{j + 2} ) is reset, and when exposure starts again, it is newly stored on the console 71 again. The above is the description of the reticle thermal deformation correction method of the present embodiment.

Next, the exposure sequence of this embodiment will be described with reference to FIGS.
This will be described with reference to FIG.

When the exposure sequence is started (Step S30), the reticle 1 is loaded on the reticle stage and chucked (Step S31). Next, using the four reticle position measurement marks (2 to 5), the reference marks (2a to 2a) provided on the reticle stage substrate 6 are used.
Reticle 1 is aligned based on 5a). The measurement coordinates of the reticle position measurement mark (2 to 5) are (x ⁽ⁱ⁾
₀ , y ⁽ⁱ⁾ ₀ ), which is the initial measurement value. next,
The first wafer 64 is transferred from the transfer system onto the wafer chuck 65 and chucked by the wafer chuck 65. Thereafter, the wafer is aligned (step S3).
3), exposure is started (step S34). While the exposure operation is being performed intermittently, the positions of the position measurement marks (2 to 5) of the reticle 1 are periodically measured, and the reticle thermal deformation is calculated (step S35). That is, the amount of thermal deformation in step S35 is determined by the deviation of the reticle position measurement mark from the initial position,

[0029]

(Equation 2) Is calculated. Next, it is determined whether or not the maximum value of the deformation amount does not exceed a predetermined value (step S36).
That is, the displacements Δx ⁽ⁱ⁾ ₁ and Δy ⁽ⁱ⁾ ₁
(I = 1 to 4) selects the maximum value of each from the line width of the target process, superimposed precalculated amount reticle thermal deformation tolerance e _x and e _y and respectively compared are from accuracy.

[0030]

(Equation 3) If the comparison result represented by the equation (3) is true, a reticle magnification correction sequence starts (step S38). If false, the exposure operation is continued unless the exposure of all wafers is completed (step S37). At this time, since the exposure has just started, the reticle is hardly thermally deformed, and it is determined that the exposure operation is to be continued. Then, the process returns to step S33 and the process for the second wafer is started. Incidentally, the operation intervals of steps S35 and S36 may be different depending on the process. For example, when the reticle pattern abundance is high, the exposure energy and the duty are large, the operation is performed for each wafer, for example. Conversely, when the reticle pattern abundance is low and the exposure energy and duty are small, the number of wafers may be five or ten. In the present embodiment, in order to constantly monitor the thermal deformation of the reticle, it is performed for each wafer.

When such a loop is repeated, the temperature of the reticle gradually increases. In the j-th mark measurement, the reticle thermal deformation amount max. Δx ⁽ⁱ⁾ _j or ma
x. [Delta] y ⁽ⁱ⁾ _j is the amount of the reticle thermal deformation tolerance e _x and e
_If it is determined that it has become larger than _y , the sequence proceeds to the reticle magnification correction sequence of step S38, and the exposure operation temporarily stops (step S42). Thermal stress due to thermal deformation is generated in the reticle 1 on which the vacuum suction is performed, and the vacuum is released to limit the thermal deformation in the x and y directions (step S43). Then, chucking is performed again, and each mark of the reticle (13 to
16 or 13a to 16a) is measured (step S44). The measurement results (Δx ⁽ⁱ⁾ _{j + 1} , Δy ⁽ⁱ⁾
_{j + 1} ), the reticle thermal deformation error,

[0032]

(Equation 4) Is calculated, and an error is assigned to each mark so that the sum of squares of the error is minimized.
And positioning in the θ direction is performed (step S45).
Reticle mark (13b-16
The measured value (x ⁽ⁱ⁾ _{j + 2} , y ⁽ⁱ⁾ _{j + 2} ) of b) is stored in the console 71. In addition, since the coordinates of the wafer alignment mark of the reticle have changed due to this alignment, the information is reflected on the wafer alignment measurement (step S46). Next, based on the coordinate values (x ⁽ⁱ⁾ _{j + 2} , y ⁽ⁱ⁾ _{j + 2} ) of the reticle position measurement mark, the generated reticle magnification is unitarily expressed in ppm,
A magnification correction value of the projection lens corresponding to the magnification component is obtained by referring to a table created in advance through experiments and simulations (step S47). In accordance with the magnification correction value, the magnification correction unit 62 of the projection lens operates to correct the magnification of the reticle (step S4).
8). If the magnification correction is ideally performed, the reticle magnification does not occur on the wafer, so the alignment measurement value of the reticle mark (13b to 16b) at this point is set as the initial value again. That is,

[0033]

(Equation 5) Then, the counter j is reset again. Then, this initial value is used as a mark initial position when each mark position of the reticle is measured when exposure is restarted.
The sequence shifts to the main sequence again, a new wafer is loaded on the stage, aligned (step S33), and the sequence of the present invention is restarted. Thus, the reticle thermal deformation amount max. Δx ⁽ⁱ⁾ _j or max. [Delta] y ⁽ⁱ⁾ _j is takes reticle magnification correction every time greater than the allowable value e _x or e _y amount reticle thermal deformation,
When all wafers to be processed have been exposed, the process ends (step S46).

The embodiment has been described above, but a case where the embodiment is applied to an actual process will be described. The correction of the reticle thermal deformation according to this embodiment is applied until the reticle reaches a thermally steady state. For example, when the reticle pattern abundance is high and the exposure energy and duty are large, the reticle temperature change is large as shown in FIG. 12, so the reticle magnification correction operation is performed at times τ10 and τ.
11, τ12, τ13. That is, immediately after the start of the exposure, the correction interval is short because the reticle suddenly warms up, and the correction interval is long near the steady state because the temperature change is small. Conversely, when the reticle pattern abundance is low and the exposure energy and duty are small, the reticle temperature change is small as shown in FIG. 13, so that only one reticle magnification correction operation needs to be performed at time τ20.

(Embodiment 2) Next, another embodiment of the present invention will be described with reference to FIGS. In this embodiment,
The feature is that the number of position measurement marks formed on the reticle is increased. In FIG. 14, reference numeral 1 denotes a reticle whose circuit pattern is not shown, and reference numerals 100 to 107 denote reticle position measurement marks whose number is increased. In this embodiment, eight marks are used. In FIG. 15, reference numeral 6 denotes a reticle stage substrate that supports a reticle stage (not shown);
Reference marks 100a to 107a provided on the substrate 6 are reference marks serving as references for the reticle position measurement marks (100 to 107). As in the first embodiment, these two types of marks face each other. In the present embodiment, by increasing the number of marks to eight, the measured value (x ⁽ⁱ⁾
_j , y ⁽ⁱ⁾ _j ) is increased, and the reticle thermal deformation amounts (Δx ⁽ⁱ⁾ _j , Δy ⁽ⁱ⁾ _j ) in step S35 are calculated,
And the maximum value of the deformation amount (max.Δ) in step S36.
x ⁽ⁱ⁾ _j , max. Δy ^{(i) _j)} a reticle thermal deformation amount permissible value (e _x, it is possible to more accurately grasp the comparison e _y). If the maximum value of the reticle deformation amount is larger than the reticle thermal deformation allowable value in step S36, the process proceeds to the reticle magnification correction sequence (step S38) in the same manner as in the first embodiment, but the reticle stage is vacuumed. Is turned off, and the reticle alignment after the thermal stress is released (step S43), the deformation error can be more accurately allocated to the eight marks (step S45). Accordingly, since the generated deformation of the reticle can be expressed more accurately as a reticle magnification, the magnification correction operation of the projection lens (step S48) can perform more appropriate correction accordingly.

The number and arrangement of the reticle position measurement marks are not limited to these embodiments, and other combinations are naturally possible depending on the circuit pattern of the reticle or the supporting position of the reticle chuck. It is.

[0037]

As described in detail above, according to the present invention,
Even if the reticle is exposed to exposure light and thermal deformation occurs, which may affect the overlay accuracy of shots, the exposure operation is temporarily stopped, and the reticle stage vacuum is turned off to generate heat. By releasing the stress, careless deformation of the reticle is eliminated. Further, by aligning the reticle again and allocating a thermal deformation error to each mark, even if asymmetric reticle thermal deformation occurs, the magnification correction function of the projection lens can be effectively operated.
By preparing a table of magnifications generated on the reticle and the corresponding magnification correction values of the projection lens on the console in advance, rapid reticle magnification correction can be performed and thermal deformation of the reticle can be corrected. I do.
As described above, a good circuit pattern is always printed on the wafer. Since the present invention utilizes a well-known magnification correction function of the projection lens, the modification of the apparatus is only to increase the number of reference marks on the reticle stage base, and the implementation is relatively simple, and does not require a large modification of the apparatus. .

[Brief description of the drawings]

FIG. 1 is a flowchart showing a main exposure sequence according to the present invention.

FIG. 2 is a flowchart showing a reticle magnification correction sequence according to the present invention.

FIG. 3 is a reticle provided with four reticle position measurement marks.

FIG. 4 is a reticle stage substrate provided with four reference marks facing reticle position measurement marks.

FIG. 5 is a diagram showing a positional relationship between a reticle and a reticle stage substrate.

FIG. 6 is a reticle thermally expanded around the center of the reticle.

FIG. 7 is an asymmetrically expanded reticle.

FIG. 8 shows the reticle of FIG. 5 which is aligned so as to allocate a deformation error to each mark after the vacuum of the reticle stage is turned off.

FIG. 9 shows a coordinate position of each mark before alignment.

FIG. 10 shows a coordinate position of each mark after alignment.

FIG. 11 is a conceptual diagram of an apparatus to which the present invention is applied.

FIG. 12 is a graph showing a transient temperature rise of a reticle having a large Cr pattern abundance.

FIG. 13 is a graph showing a transient temperature rise of a reticle having a small Cr pattern abundance.

FIG. 14 is a reticle according to a second embodiment.

FIG. 15 shows a reticle stage substrate according to a second embodiment.

[Explanation of symbols]

1: reticle, 6: reticle stage substrate, 2 to 5: reticle position measurement mark, 2a to 5a: reference mark, 6
1: Projection lens, 62: Projection lens magnification correction function.

Claims (4)

[Claims] 1. A projection exposure apparatus for illuminating a circuit pattern on an original plate with illumination light and projecting and exposing the same onto a substrate to be exposed via a projection lens, wherein a heat generated by the original plate absorbing the illumination light is provided. In order to detect deformation, four or more position measurement marks are provided on the original plate, and at the same time, four or more reference marks are provided on the base board supporting the original plate so as to face the position measurement marks. Projection exposure means for measuring a deviation of a coordinate value of the position measurement mark with respect to each of the reference marks and correcting a change in a pattern projected on the substrate to be exposed in consideration of the deviation. apparatus. 2. When the measured amount of thermal deformation of the original plate exceeds a predetermined allowable value, the projection exposure operation is temporarily stopped, and the stage is supported on the original plate and is movable parallel to the base plate. After releasing the function of fixing the original plate and releasing the thermal stress, the original plate is fixed to the stage again, and the sum of squares of the position errors of the position measurement marks caused by the thermal deformation thereof. 2. The projection exposure apparatus according to claim 1, further comprising means for aligning the stage so that the distance is minimized. 3. A magnification device generated on the original plate and a correction amount of a magnification correction function of the projection lens corresponding to the magnification component are stored in a storage device in advance as a table, and coordinate values of the position measurement marks are stored. 3. The projection system according to claim 1, further comprising: means for converting a magnification component due to thermal deformation of the original plate, and performing magnification correction by obtaining a correction amount of the magnification correction function corresponding to the converted value from the table. Exposure equipment. 4. The projection exposure according to claim 3, wherein, after the magnification correction of the original plate is performed, the coordinate value of each of the position measurement marks of the original plate is set as the initial coordinate of the reticle in the restarted exposure operation. apparatus.