JP2003303765A - Measurement method of amount of distortion aberration and image surface curvature - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To prevent an influence of the pattern shape of a reference pattern, to enable the accurate measurement of distortion aberration and image surface curvatures, and to carry out the separation and correction of the amount of distortion aberration caused by the arrangement error of a reference reticle and a mask and that in a projection optical system single unit. <P>SOLUTION: The mask 6 is reciprocatingly scanned in a direction vertical relative to an optical axis for periodic motion, thus detecting the phase difference between time distribution in the quantity of light that is measured at an opening 60 formed on the optical axis of the projection optical system 5 and time distribution in the quantity of light that is measured at each opening 61 provided outside the optical axis of the projection optical system 5. The phase difference enables the measurement of the distortion aberration and the image surface curvature. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of measuring the amount of distortion in a projection optical system such as a projection exposure apparatus for manufacturing semiconductor devices.

[0002]

2. Description of the Related Art A conventional method for measuring distortion and field curvature will be described with reference to FIG.

For example, the distortion and field curvature of the projection optical system have been measured as follows. That is, light emitted from a light source (not shown) similarly passes through an illumination optical system (not shown), and a plurality of light beams having the same line width on the reference reticle 102 installed on the object-side focal plane of the projection optical system 101. The reference pattern 103 is illuminated and an image is formed at a position conjugate with the reference pattern 103 via the projection optical system 101.

On the other hand, each reference pattern 1 is formed at this image formation point.
03, a mask 107 having a plurality of openings 104 of the same shape is movably provided.
The image of the reference pattern 103 incident on
Is converted into an electric signal according to the amount of light by the photo-detecting element 105 provided in the computer and sent to the computer 106 through an interface (not shown). The computer 106 controls each unit (not shown) necessary for calculation and measurement of the distortion aberration and the field curvature from the above electric signal.

The distortion aberration is measured by scanning the mask stage 108, on which the mask 107 is mounted, in one direction on a plane perpendicular to the optical axis. At this time, the transmitted light amount distribution of the reference pattern 103 received by the light detection element 105 on the side of each opening 104 is as shown in FIG.

That is, the time when the central aperture 109 provided on the optical axis of the projection optical system 101 shown in FIG. 9 coincides with the center of the transmitted light amount distribution, and the aperture 104 provided outside the optical axis, that is, (i, When the time difference from the time when the mask in j) coincides with the center of the transmitted light amount distribution is δt (i, j),
Distortion aberration amount dx (i, j) in the scanning direction of the mask stage 108 at the opening 104 outside the optical axis, that is, (i, j)
Is calculated by the following equation: dx (i, j) = v × δt (i, j) (where v represents the scanning speed of the mask stage 108).

Next, when scanning is performed in the y direction orthogonal to the first scanning direction (for example, the x direction), the distortion aberration amounts dx (i, j), dy in the two directions of x and y of the projection optical system 101.
It is possible to measure (i, j).

Further, in measuring the field curvature, the stage 805 is scanned in one direction along the optical axis to obtain a transmitted light amount distribution. When the time difference between the center of the transmitted light amount distribution in the aperture provided on the optical axis of the projection optical system in FIG. 8 and the center of the transmitted light amount distribution in the off-axis aperture (i, j) in FIG. 8 is δt (i, j) , The field curvature amount dz (i, at the off-axis aperture (i, j)
j) is calculated by dz (i, j) = v × δt (i, j).

The central position of the transmitted light amount distribution is calculated by the following method.

Normally, the pulse width of the excimer laser used as the exposure light source is sufficiently short with respect to the scanning time of the mask 107, so that the light amount signal is measured as a discrete distribution as shown in FIG. This discrete k-th signal is I
k, and tk is the time at which the signal Ik is measured, the time t when the center of the light amount distribution passes through the aperture (i, j).
(I, j) is calculated by the following equation: t (i, j) = (Σtk ⁻ × Ik) / (ΣIk). That is, the position of the center of gravity of the light quantity was calculated as the center of the light quantity distribution.

[0011]

However, there is a phenomenon that the amount of distortion aberration and the curvature of field of the projection optical system change depending on the shape of the printing pattern of interest. Therefore, in order to measure the strict distortion amount and the field curvature in the projection optical system, it is necessary to measure the distortion amount and the field curvature in a plurality of pattern shapes.

By the way, in the conventional measurement of the distortion aberration and the field curvature by the calculation of the center of gravity of the light quantity, there is a problem that the measurement accuracy is affected by the relationship between the reference pattern shape and the aperture width. Especially, the influence is remarkable in the pulse light source which is the mainstream as the exposure light source at present. For this reason, it is difficult to measure the distortion aberration amount and the field curvature of the projection optical system with high accuracy regardless of the printing pattern shape of interest, and the light is transmitted through the aperture in order to improve the measurement accuracy. It is necessary to increase the number of transmitted light amount data regarding light.

As a countermeasure therefor, for example, the scanning speed of the mask may be reduced or the width of the opening may be widened. However, in the former case, there is a drawback that the time required for measurement becomes long, and in the latter case. Had the drawback of lowering measurement accuracy. In addition, since the measurement value of the distortion aberration amount includes a measurement error due to the reference pattern and the arrangement error of the aperture, it is difficult to measure the distortion aberration amount and the field curvature with a single projection optical system. It also had the drawback of being present.

Therefore, in view of the above-mentioned circumstances, the present invention is not easily affected by the pattern shape of the reference pattern, enables high-precision measurement of distortion and field curvature, and disposition error of the reference reticle and mask. An object of the present invention is to provide a method of measuring a distortion aberration amount that can perform separation and correction of the distortion aberration amount caused by the above and the distortion aberration amount of the projection optical system alone.

[0015]

In order to solve the above problems, the invention according to claim 1 provides a reference reticle having a plurality of reference patterns formed on the object plane of a projection optical system, and a projection reticle. A mask having a plurality of openings on the image plane of the optical system, a photodetector for measuring the amount of light transmitted through the openings, and a moving mechanism for moving the mask in a direction perpendicular to the optical axis and in the optical axis direction. In the method of measuring the amount of distortion and field curvature of the projection optical system from the output of the time distribution of the amount of light of the photodetector, the method of measuring the amount of distortion and field curvature, the mask in the direction perpendicular to the optical axis. And reciprocally scanning in the direction of the optical axis to perform periodic motion, and temporal distribution of the amount of light measured at the aperture provided on the optical axis of the projection optical system and individual elements provided outside the optical axis of the projection optical system. Distribution of light intensity measured at the aperture of Detecting a phase difference, it is characterized in that the measurement of distortion and field curvature by the phase difference.

According to a second aspect of the present invention, in the invention according to the first aspect, when the mask is moved cyclically, a temporal distribution of the amount of received light at an aperture provided on the optical axis of the projection optical system and The phase difference between the temporal distribution of the amount of light measured in each aperture provided outside the optical axis of the projection optical system,
Using the frequency distribution of the amount of received light obtained by expanding the time distribution of the amount of received light in each aperture, it is possible to calculate as the difference between the center intervals of the time distribution of the amount of received light in the forward path and the return path during the reciprocating periodic movement of the mask. It is a feature.

According to a third aspect of the invention, in the invention according to the first or second aspect, the reference pattern provided on the reference reticle corresponding to the plurality of openings has a plurality of reference patterns for each opening. The present invention is characterized in that patterns having different shapes are measured, and distortion aberration of the projection optical system with respect to a predetermined reference pattern among the plurality of reference patterns having different shapes is measured.

The invention according to claim 4 is a reference reticle having a plurality of reference patterns formed on the object plane of the projection optical system, and a mask having a plurality of openings on the object plane of the projection optical system. A photodetector for measuring the amount of light passing through the opening, and a moving mechanism for moving the mask in parallel with the optical axis in a direction perpendicular to the optical axis. In the distortion aberration amount measuring method for measuring the distortion aberration amount, the measurement value of the distortion aberration in a state where the reference reticle and the mask are rotated and laterally displaced with respect to the projection optical system, and the measurement values of the reference pattern interval and the aperture interval. Therefore, the distortion aberration caused by the pattern arrangement error of the reference reticle and the mask is separated from the measured distortion aberration amount and corrected.

[0019]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the accompanying drawings.

FIG. 1 shows a measuring apparatus used in the method for measuring the amount of distortion and the amount of curvature of field according to the embodiment of the present invention.

The apparatus for measuring the amount of distortion and the amount of curvature of field includes a light source 1, an illumination optical system 2, a reticle stage 4 having a reference reticle 3 mounted thereon, a light projecting optical system 5, and a mask 6.
The mask stage 7 on which the wafer is mounted, the light amount detector 8, the wafer stage 10 on which the wafer 9 is mounted, and the controller 11. And in this distortion aberration amount measuring device,
The light emitted from the light source 1 passes through the illumination optical system 2 and illuminates the reference reticle 3 installed on the object-side focal plane of the projection optical system 5, and the reference pattern 31 on the reference reticle 3 is illuminated. Form an image at a conjugate position via the projection optical system 5.

As shown in FIG. 2A, the reference reticle 3 has a large number of reference patterns 31 arranged in a grid pattern.
In the reference pattern 31, as shown in an enlarged view of one square in the reference reticle 3 in FIG. 6B, a black portion indicates a transmission area B where the illumination light is transmitted. In the reference pattern 31 of this embodiment, the transmissive area B is formed by the pattern shape of four isolated lines having different line widths W1 to W4, but this is a so-called line & space (L & S) pattern. I do not care.

In this reference pattern 31, the interval Δ1 (in the scanning direction) between the patterns having a plurality of line widths W1 to W4 provided corresponding to each opening 61 described later is the reference pattern 31 of the shape of interest. In order to prevent the signal of the image of the reference pattern 31 of other shape from being mixed in the measurement of the distortion aberration in (1), the image side conversion value of the scanning amount of the mask stage 7 (the scanning amount is δ, the projection optical system 5 If the magnification of is β, it must be larger than | βδ |). Here, the scanning amount δ requires a value equal to or larger than the distortion aberration of the projection optical system 5 to be measured. In order to avoid that the pattern used for the scanning in the direction orthogonal to the scanning direction is mixed by the distortion aberration of the projection optical system 5 when the mask 6 is scanned, the length Δ2 of each pattern is also longer than the above | βδ |. There is a need to. At the time of measuring the field curvature, in order to prevent the signal of the image of the reference pattern 31 other than the shape of interest from being mixed, Δ1
Is the object-side equivalent of the expected distortion of the measurement optical system (where d is the amount of distortion and β is the magnification of the projection optical system, | d /
Being larger than β |) is required. Further, in order to make the image of the reference pattern 31 incident on the entire mask 6 regardless of the distortion, the length Δ2 of each pattern also needs to be longer than | d / β |.

As shown in FIG. 3, the mask 6 has a large number of openings 61 shown in white, which are regularly arranged in a grid pattern. When the distortion of the projection optical system 5 is zero, the center of each of the openings 61 is the reference pattern 31.
Are arranged so as to have a conjugate relationship with the respective centers of the projection optical system 5. That is, the center position of the arbitrary reference pattern 31 on the reference reticle 3 and the center position of the opening on the mask 6 corresponding to the center position of the reference pattern 31, for example, in FIG.
Letting ro (i, j) and ri (i, j) be the ideal imaging magnification of the projection optical system 5 is β · ro (i, j) = ri (i, j).

Here, in order to measure the distortion aberration caused by the device which may be added to the distortion aberration of the projection optical system 5 which should be originally obtained, the reference reticle 3 is arranged on the optical axis of the projection optical system 5. With respect to the reference pattern 30 (hereinafter referred to as the central reference pattern), the opening 60 (hereinafter referred to as the central opening) arranged on the optical axis of the projection optical system 5 in the mask 6 is the central opening. It must have rotational symmetry about 60 and translational symmetry in the x, y directions.

The photo-detecting element 8 is provided corresponding to the opening 61 of the mask 6, and converts the amount of light transmitted through each opening 61 into an electric signal corresponding to the amount of transmitted light, thereby simultaneously transmitting the amount of transmitted light at a plurality of image points. Can be calculated. The light detecting element 8 may be directly installed on the mask 61, or may be installed outside by propagating the amount of transmitted light with a fiber or the like. The photo-detecting element 8 is connected to the computer 11 through an interface in order to perform the calculation process of the output electric signal. The computer 11 has a function of sending a command necessary for a series of distortion aberration measurement such as control of the mask stage 7 to each unit (not shown) as well as a calculation process associated with the calculation of the aberration amount.

The mask stage 7 is installed on the wafer stage 10 on which the wafer 9 is mounted, and the wafer stage 10 drives the wafer 9 and the mask 6 when the wafer 9 is exposed and the distortion aberration is measured. The mask stage 7 includes a moving stage 71 that moves by an x, y direction moving mechanism (not shown), a moving stage 72 that moves by a z direction moving mechanism (not shown), and a θ direction moving mechanism (not shown). And a rotary stage 73 that rotates by. In addition, regarding movement in the x, y, and z directions, in particular, a piezoelectric element is used for use in scanning when measuring distortion and field curvature, and by applying feedback to this with an interferometer, precise It enables high speed driving. Further, the moving mechanism in the θ direction includes the reference pattern 31 and the opening 6.
This is for measuring the distortion aberration caused by the placement error of 1. Regarding the reference reticle stage 4,
A step drive corresponding to the interval of the reference pattern 31 in the x and y directions and a 90-degree rotation mechanism are provided. Regarding this rotation, the stage may not be provided with a rotation mechanism, and after the one-way measurement is completed, the reference reticle 3 may be temporarily removed, and then rotated and reattached to the projection exposure apparatus.

Example 1 The procedure for measuring the distortion aberration in this embodiment will be described below as the first example.

First, the reference reticle 3 is aligned so that the optical axis of the projection optical system 5 passes through a predetermined pattern shape of the central reference pattern 30 of the reference reticle 3, and then the wafer stage 10 is driven to drive the mask 6 Is moved to the object-side image plane of the projection optical system 5. As a result, the light incident on the image of the central reference pattern 30 passes through the central opening 60 of the mask 6 and the photodetector 8 provided corresponding to the central opening 60.
Adjust to receive light.

Next, the mask stage is adjusted in the z direction. This is so-called focusing. As a method for this focusing, the mask stage 7 is stepped in the z direction, and the rise or rise of the transmitted light amount signal of the image of the reference pattern 30 when the mask 6 is scanned in the x direction or the y direction at each step. The contrast may be calculated from the sharpness of the fall. To calculate the contrast, a differential value of the measured light amount signal with respect to time may be used. After the contrast measurement in a plurality of steps is completed, the mask stage 72 is moved to the z coordinate position where the contrast is the best.
To drive. Further, as another focusing method, a focusing position adjusting mechanism provided in the exposure apparatus main body may be used.

After the focus position adjustment, reciprocal scanning in the x direction is restarted, and the central position of the transmitted light amount position at the central aperture 60 on the optical axis is calculated. The method of calculating the center position will be described later. After the calculation of the center position, the wafer stage 10 is driven and the mask stage 7 is finely adjusted so that the transmitted light amount of the central opening 60 on the optical axis is near the center position of the reciprocating scanning of the mask stage 7. The fine adjustment may be performed by the voltage of the piezoelectric element of the mask stage 71. Adjustment in the y direction is performed in the same manner. In addition,
The central position of the transmitted light amount of the central aperture 60 on the optical axis is aligned with the center of the reciprocal scanning of the mask stage 7 when the distortion of the projection optical system 5 has a large value. This is to avoid that the scanning of the mask stage 7 becomes insufficient and the distortion aberration cannot be measured.

When the above adjustment is completed, the measurement is started. Distortion aberration is obtained by measuring the components in each direction by separately performing x and y in the scanning direction, and then synthesizing them.

Next, a method of measuring the distortion aberration dx in the x direction will be concretely described.

The amount of transmitted light obtained when the mask stage 71 is reciprocally scanned in the x direction is shown in FIG. Here, the reciprocal scanning needs to perform highly accurate constant velocity scanning in a scanning region having the same amount of distortion as the measurement target, which can be performed by the above-described piezoelectric element stage with feedback. In Fig. 4, the outside signal (solid line part)
Is a light amount signal of the opening 61 outside the optical axis when distortion is generated, and the signal inside (the wavy line portion) is the opening 6 on the optical axis.
It is a light amount signal at 0. This is because the difference in the image point position of the reference pattern 31 due to the distortion aberration dx causes a difference in the intervals of the amounts of transmitted light measured in the forward path and the backward path.

Therefore, the time interval (time difference) δt of the central position of the light quantity measured at any aperture 61, for example, the aperture of M (i, j), during the scanning in the forward and backward passes.
(I, j) and the central aperture 60 on the optical axis, that is, M (0,
0), the difference between the central position of the light quantity measured at each aperture (time difference) δt (0, 0) and δT (i,
j), that is, δT (i, j) = δt (i, j) −δt (0,0) (0), where dx (i, j) = where v is the scanning speed during constant-speed scanning. The distortion aberration dx in the x direction can be obtained as v · δT (i, j) / 2 (1). The scanning speed of the mask stage 71 is not constant at the time of folding in the scanning direction, but since the influence thereof evenly spreads over all the openings 60 and 61, it does not affect the measurement of distortion at all. Further, for the signals of FIG. 4, assuming the exposure of the wafer 9, and considering the influence of the photosensitive material, a predetermined reference intensity is set for the intensity of the signals, and only signals having an intensity higher than that are provided. You can use it.

Here, from the signal of FIG. 4, the distribution of the received light amount of the aperture 61 provided outside the optical axis of the projection optical system 5 showing the distortion and the received light amount distribution of the central aperture 60 provided on the optical axis. A method of obtaining the phase difference from the difference between the time intervals (time difference) δt of the central positions of the time distributions of the light amounts on the forward and backward paths will be described.

First, a conversion value (hereinafter referred to as a spectrum) S1 to the frequency domain is calculated from the light amount distribution in the time domain using the entire signal of FIG. 4 (a signal including both the light intensity distributions in the forward path and the return path). As a method of calculating this spectrum, a high-speed Fourier transform (hereinafter referred to as FFT)
The maximum entropy method or the like may be used. In this embodiment, FFT is used.

Next, similarly, the spectrum S2 is calculated for the signal including only one side of the light amount distribution of FIG. The light intensity distribution on one side may use either the forward or backward path, or both spectra may be calculated and the average of the two may be taken as S2.

Here, an example of calculating a spectrum from the light quantity distribution shown in FIG. 5 will be shown. As can be seen from FIG. 5, S1 has S2 in the envelope, and shows fine vibrations depending on the frequency corresponding to the interval between the center positions of the light amount distributions on the outward path and the return path. Therefore,
If the frequency interval δf (i, j) of the fine vibration of S1, that is, the phase difference is obtained, δt (i, j) = 1 / δf (i, j) (2) It is possible to obtain the time interval (time difference) δt between the light quantity distribution centers of the forward and backward paths.

In order to accurately calculate δf (i, j), it is necessary to remove the S2 component from the S1 component.
For that purpose, S1 should be standardized by S2, and the result is as shown in FIG. The spectrum of FIG. 6 is designated as S3. In order to obtain the vibration frequency interval from S3, the frequency value of the peak next to the center (DC component) may be calculated. Also,
As another method, the average of the frequency intervals of the peaks may be used, or S3 may be partially cut out and the FFT may be performed again. In this embodiment, the isolated pattern shown in FIG. 3 is used as a reference, but in the case of the L & S pattern, only the shape of S2 changes, and δf (i, j).
The calculation method of is the same.

Here, simply converting the measured signal as it is requires the spectral resolution δf = 1 / (δts × n) (3) determined by the sampling interval δts and the sampling number n shown in FIG. Since it does not meet the required accuracy, it is necessary to add a zero point to the end of the measurement signal so that the required frequency resolution δf can be obtained. In this embodiment, it is necessary to add tens of thousands to hundreds of thousands of sampling numbers. As a result, it is necessary to execute the FFT for an enormous number of data, but since the focus is only on the vicinity of the frequency corresponding to the half cycle of the reciprocating motion of the mask stage 7, 1 / δ
It is not necessary to calculate the spectrum over the entire frequency band defined as ts, and the zero can be deleted from the loop in the calculation algorithm, so that the substantial calculation can be performed at high speed.

After the above calculation is performed on the light amount distribution in the entire aperture 61, the light amount distribution time difference δt (0,0) in the central aperture 60 on the optical axis is subtracted,
ΔT (i, j) is calculated from the equation (0), and the amount of distortion aberration is calculated from the equation (1). This completes the measurement of the distortion aberration in the x direction. The distortion aberration in the y direction is the same as that in the x direction except that the scanning direction of the mask stage 71 is changed. As for the distortion aberration calculated as described above, the amount of distortion aberration as a statistic such as a symmetric component may be calculated if necessary. Further, if reciprocal scanning is repeated as needed, the influence of errors due to random factors such as variations in scanning speed can be suppressed, and more accurate measurement can be performed.

Thus, the measurement of the distortion of the projection optical system 5 for one line width of the reference pattern 31 is completed. When changing the pattern shape of interest, the mask 6 is moved by driving the wafer stage 7, the mask 6 is moved to the image point position of the pattern shape of interest, and the same procedure is performed again.

As described above, a method of calculating the center position interval of the light quantity distributions on the outward path and the return path from the spectrum and measuring the phase difference between the light quantity distributions at the central openings on the optical axis and the light quantity distributions at the openings outside the optical axis is provided. FIG. 7 shows a comparison of the measurement accuracy between the distortion aberration measurement value when used and the conventional distortion aberration measurement value due to the light amount centroid.

FIG. 7 shows the reproducibility of distortion by simulation. Here, the energy variation of the exposure light source 1 was set to σ = 2.5%. Assuming that isolated lines with different line widths are used as patterns of different shapes, the reproducibility in 100 repeated measurements is 3 with respect to the ratio with the opening width.
Compared by σ. According to FIG. 7, according to the method of measuring the distortion aberration calculated from the spectrum, the accuracy does not change with respect to the shape of the reference pattern 31, and the highly accurate measurement is possible, as compared with the conventional method of calculating from the light amount centroid. It turns out that it will be possible.

Next, a method of correcting the distortion aberration due to the manufacturing error of the reference reticle 3 and the mask 6 will be described. In addition,
If the reference pattern 31 and the opening 61 have a placement error,
The measured value of the distortion aberration is added to the distortion aberration of the projection optical system 5 alone, so that the distortion error of the projection optical system 5 needs to be separated in order to measure the distortion aberration of the projection optical system 5. . (I) Therefore, first, a method of measuring the distortion aberration caused by the projection optical system, such as the manufacturing error of the reference pattern 31 of the reference reticle 3 and the opening 61 of the mask 6, will be described.

First, with respect to both the reference pattern 31 and the opening 61, the positions of the reference reticle 3 and the mask 6 are adjusted so that the central opening 60 and the central reference pattern 30 are aligned on the optical axis of the projection optical system 5, as in the normal measurement. Make adjustments, x and y
The distortion aberration DR0 (i, j) is measured by scanning in the direction.

Next, the mask is shifted in the x direction by the aperture interval, and the distortion aberration DR1x (i, j) is measured. At this time, the distortions measured at the respective apertures (i, j) have the following relationships.

DR0 (i, j) = dr0 (i, j) [Lens + Reticle] + dr (i, j) [Mask] (4) DR1x (i, j) = dr1x (i, j) [Lens + Reticle] + dr (I, j) [Mask] -dr (1,0) [Mask] (5) Here, dr (i, j) [Lens + Reticle] is
This is distortion aberration caused by the projection optical system 5 and the reference reticle 3, and this measurement value does not change even if the measurement is performed by shifting the aperture 61. Therefore, dr1x (i, j) = dr0.
There is a relationship of (i + 1, j).

On the other hand, dr (i, j) [Mask] is a distortion aberration caused by the mask 6 with the central aperture 60 on the optical axis as a reference, and this is different for each aperture 61 used for measurement. The measurement value of the distortion aberration at the same image point of the projection optical system 5 changes depending on the aperture used for the measurement. (4), (5)
From the formula, dr (i, j) [Mask] = DR0 (i, j) -DR1x (i-1, j) + dr (i-1, j) [Mask] -dr (1,0) [Mask] ... ... (6) is obtained. Since this equation (6) constitutes the form of the recurrence equation for i of dr (i, j) [Mask], i =
When expanded to 0, when i> 0, dr (i, j) [Mask] = DR0 (i, j) + DR0 (i-1, j) + ...- DR1x (i-1, j) -DR1x (i −2, j) −...− | i | dr (1,0) [Mask] + dr (0, j) [Mask] (7) is obtained.

Similarly, the measured value DR1y in the case where the distortion aberration is measured by shifting the mask 6 in the y direction by the aperture interval.
From the comparison with (i, j), if j> 0, then dr (i, j) [Mask] = DR0 (i, j) + DR0 (i, j-1) + ...- DR1y (i, j-1) ) -DR1y (i, j-2) -...- | j | dr (0,1) [Mask] + dr (i, 0) [Mask] ... (8).

For negative i or j, (7), (8)
The aperture coordinate values such as i-1, j-1 in the formula are i + 1, j + 1.
As a result, i and j may be reduced to 0. If i = 0 in the equation (8), as the final term of the equation (7), dr (0, j) [Mask] = DR0 (0, j) + DR0 (0, j-1) + ...- DR1y (0, j-1) -DR1y (0, j-2) -...- | j | dr (0,1) [Mask] ... (9). Substituting this into equation (7), dr (i, j) [Mask] = DR0 (i, j) + DR0 (i-1, j) + ...- DR1x (i-1, j) -DR1x (i -2, j) -...- | i | dr (1,0) [Mask] + DR0 (0, j) + DR0 (0, j-1) + ...- DR1y (0, j-1) -DR1y (0, j-2) -...- | j | dr (0,1) [Mask] (10). Therefore, the amount of distortion aberration dr (i, j) [Mask] due to the aperture at an arbitrary aperture (i, j) is determined by the lattice spacing error dr (1,0) [Mask] in the x direction and the lattice spacing error in the y direction. It can be expressed by an equation in which only dr (0,1) [Mask] is an unknown number. (II) Next, the mask stage 73 is rotated by 90 degrees to measure distortion. The opening 61 with the mask 6 in the 0 degree state,
For example, focusing only on two measurement values at M (0,1) and M (1,0), which are the same image points in the 90 ° rotation state, the distortion aberration measurement values at the 0 ° and 90 ° states are DR0
(0,1), DR90 (1,0), DR0 (0,1) = dr0 (0,1) [Lens + Reticle] + dr0 (0,1) [Mask] (11) DR90 (1,0) = Dr90 (1,0) [Lens + Reticle] + dr90 (1,0) [Mask] (12) Since the distortion aberration caused by the projection optical system 5 and the reference reticle 3 does not depend on the rotation of the mask 6, dr0 (0,1) [Lens + Reticle] = dr90 (1,0) [Lens + Reticle] (13) , (11) to (13), dr0 (0,1) [Mask] = dr90 (1,0) [Mask] + DR0 (0,1) -DR90 (1,0) = Rotate [dr0 (1,0) 0) [Mask], 90] + DR0 (0,1) -DR90 (1,0) (14) (where Rotate [dr, θ] numerically indicates that dr is rotated by θ. ) Is obtained. Thereby, from the results of (10) and (14), the distortion aberration dr0 (i, j) [Mask] caused by the mask 6 is dr0.
Only (1,0) [Mask] can be represented as an unknown number. Here, dr0 (1,0) is a placement error of the aperture 61, that is, M (1,0) with reference to the central aperture 61 on the optical axis, and the placement of other gratings is compensated with this value as a reference. Will be done. That is, the placement error of the mask 6 in the form of a lattice is compensated, and the remaining gaps of the lattice | dr0 (1,0) | and the lattice rotation deviation dr0 (1,
0) [Mask] / | dr0 (1,0) [Mask] | (III) Finally, the distance between the lattices, that is, the distance between the openings is measured.

First, the central opening 60 of the mask 6, that is, M
(0, 0) is moved on the optical axis of the projection optical system 5, the center of the light amount is calculated, and the position is set as the origin of the interferometer.

Next, the mask 6 is moved by the lattice spacing, the opening 61, for example, M (1,0) is arranged on the optical axis of the projection optical system 5, and the center of light quantity is calculated in the same manner. Let L be the coordinate value of the interferometer, which corresponds to the center position of the light amount distribution of M (1,0), in this opening 61. dr0 (1,0) [Mask] = Lm-Lm0 (15) (here) Where Lm0 is the design value of the lattice spacing.) From the above three steps, that is, the distortion aberration measurement values according to (I) to (III), from the relationship of equations (10), (14), and (15), dr ( i, j) [Mask] ^- = DR0 (i, j) + DR0 (i-1, j) + ...- DR1x (i-1, j) -DR1x (i-2, j) -... + DR0 (0, j) ) + DR0 (0, j-1) + ...- DR1y (0, j-1) -DR1y (0, j-2) -...- | i | * (Lm-Lm0)-| j | * (Rotate [( Lm−Lm0), 90] + DR0 (0,1) −DR90 (1,0)) (16) To calculate the amount of distortion caused by the 6.

The reference pattern 31 of the reference reticle 3 is also calculated by using the same mechanism as the mask 6 and performing the same method. At this time, the distortion aberration resulting from the arrangement error of the reference pattern 31 is dr (i, j) [Reticle] = DR0 '(i, j) + DR0' (i-1, j) + ...- DR1x '(i- , J) -DR1x '(i-2, j) -... + DR0' (0, j) + DR0 '(0, j-1) + ...- DR1y' (0, j-1) -DR1y '(0, j-2) -...- | i | * (Lr-Lr0)-| j | * (Rotate [(Lr-Lr0), 90] + DR0 '(0,1) -DR90' (1,0)) ... (17) (where Lr is the measured reference pattern interval, Lr0
Represents the design value of the reference pattern interval, and 'represents the measured value of the distortion aberration when the reference reticle 3 is moved similarly to the mask 6. )It can be expressed as.

As described above, from the equations (16) and (17), the distortion aberration DR (i, j) [lens] of the projection optical system 5 alone.
Can be expressed by the following equation: DR (i, j) [lens] = DR0 (i, j) -dr (i, j) [Mask] -dr (i, j) [Reticle] (18) .

Therefore, dr (i, j) [Mask] and dr
Once (i, j) [Reticle] is calculated,
By correcting the measurement value using the equation (18), the net distortion of the projection exposure system 5 can be obtained.

Example 2 The procedure of measuring the field curvature in this embodiment will be described below as a second example.

First, the reference reticle 3 is aligned so that the optical axis of the projection optical system 5 passes through a predetermined pattern shape of the central reference pattern 30 of the reference reticle 3, and then the wafer stage 10 is driven to move the mask 6 Is moved to the object-side image plane of the projection optical system 5. As a result, the light incident on the image of the central reference pattern 30 passes through the central opening 60 of the mask 6 and the photodetector 8 provided corresponding to the central opening 60.
Adjust to receive light.

Next, the mask stage is adjusted in the x, y and z directions.

The adjustment in the z direction is so-called focusing. As a method, the transmitted light amount distribution in the reference mask 60 when the mask 6 is reciprocally scanned in the z direction is used. The center of the transmitted light amount is adjusted to the center position of reciprocal scanning. In particular,
While performing reciprocal scanning, gradually move the center position of the reciprocation to 1
It suffices to set the center of the reciprocating motion so that the frequency component that is an odd multiple of the mask drive in the transmitted light amount distribution when scanning in the direction is minimized. This is based on the principle that, when the center position of the reciprocal scanning of the mask 6 and the center position of the transmitted light amount distribution coincide with each other, the spectrum of the transmitted light amount distribution has only frequency components that are an even multiple of the mask drive. is there. The same applies to adjustments in the x and y directions. As another focusing method, the focal position may be determined from the change in contrast described above, or a focal position adjusting mechanism provided in the exposure apparatus main body may be used.

When the above adjustment is completed, the measurement is started. The field curvature is obtained by performing the scanning direction in z.

FIG. 8 shows the amount of transmitted light obtained when the mask stage 71 is reciprocally scanned in the z direction. Figure 8
The middle and outer signals (solid line portions) are light amount signals of the aperture 61 outside the optical axis when the field curvature is generated, and the inner signals (wavy line portions) are light amount signals at the aperture 60 on the optical axis. is there.
This is because a difference in the image point position of the reference pattern 31 due to the amount of curvature of field dz causes a difference in the interval at which the amount of transmitted light measured on the outward path and the return path becomes maximum.

Therefore, the time interval (time difference) δt of the maximum position of the light quantity measured at any aperture 61, for example, the aperture of M (i, j), during the scanning in the forward pass and the backward pass.
(I, j) and the central aperture 60 on the optical axis, that is, M (0,
0), the difference between the central position of the light quantity measured at each aperture and the time interval (time difference) δt (0,0) is δT (i, j).
Then, it is possible to obtain the field curvature amount dz by setting dz (i, j) = v.delta.T (i, j) / 2 (19), where v is the scanning speed during constant-speed scanning. Further, for the signals of FIG. 4, assuming the exposure of the wafer 9, and considering the influence of the photosensitive material, a predetermined reference intensity is set for the intensity of the signals, and only signals having an intensity higher than that are provided. You can use it.

From the signal of FIG. 8, the phase difference between the received light amount distribution of the aperture 61 provided outside the optical axis of the projection optical system 5 representing the field curvature and the received light amount distribution of the central aperture 60 provided on the optical axis. The method of obtaining the difference from the time interval (time difference) δt of the central positions of the time distributions of the forward and backward light amounts is the same as that in the measurement of the distortion aberration amount of the first embodiment, and therefore will be omitted.

After the above calculation is performed on the light amount distribution in the entire aperture 61, the light amount distribution time difference δt (0,0) in the central aperture 60 on the optical axis is subtracted to obtain
ΔT (i, j) is calculated from the equation (0), and the curvature of field is calculated from the equation (19). With the above, the measurement of the field curvature is completed. By repeating the reciprocal scanning as necessary, it is possible to suppress the influence of errors due to random factors such as variations in the scanning speed and to perform more accurate measurement.

Thus, the measurement of the field curvature of the projection optical system 5 for one pattern shape of the reference pattern 31 is completed.
When changing the pattern shape of interest, the mask 6 is moved by driving the wafer stage 7, the mask 6 is moved to the image point position of the pattern shape of interest, and the same procedure is performed again.

[0068]

As described above, according to the present invention, the amount of distortion is calculated from the spectrum of the received light amount distribution from the phase difference of the received light amount distribution between the aperture outside the optical axis and the aperture on the optical axis. By suppressing the influence of the pattern shape of interest, it becomes possible to measure the distortion amount and the field curvature of the projection optical system with high accuracy and high speed.

Further, according to the present invention, the measured value of the distortion aberration in the rotation state with the reference pattern on the optical axis of the projection optical system and the aperture as the rotation center and the measurement of the distortion aberration when the reference reticle and the mask are laterally displaced. From the values and the measured values of the reference pattern spacing and the aperture spacing, it becomes possible to separate and correct the distortion aberration amount caused by the placement error of the reference reticle and the mask from the distortion aberration amount of the projection optical system alone.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram showing an embodiment of the present invention.

2A is an explanatory view showing a reference reticle according to an embodiment of the present invention, and FIG. 2B is an enlarged view showing the reference pattern.

FIG. 3 is an explanatory view showing an opening of the mask according to the embodiment of the present invention.

FIG. 4 is a correlation diagram showing a relationship between time and the amount of light obtained on and off the optical axis when the mask stage is scanned during distortion measurement.

FIG. 5 is a graph showing spectra S1 and S2 of a light amount signal.

FIG. 6 is a graph showing spectra S2 and S3 of a light amount signal.

FIG. 7 is a graph showing a correlation between a line width-aperture width ratio and distortion reproducibility.

FIG. 8 is a correlation diagram showing a relationship between time and the amount of light obtained on and off the optical axis when the mask stage is scanned during field curvature measurement.

FIG. 9 is a schematic configuration diagram showing a conventional example.

FIG. 10 is a correlation diagram showing the relationship between the amount of light obtained on and off the optical axis when scanning the conventional mask stage and time.

[Explanation of symbols]

1 exposure light source 2 Illumination optical system 3 standard reticle 30 central reference pattern 31 Reference pattern 4 Standard reticle stage 5 Projection optical system 6 masks 60 central opening 61 opening 7 Mask stage 71 Moving Stage 72 Moving stage 73 rotation stage 8 Photodetector (photodetector) 9 wafers 10 Wafer stage 11 computer dr Distortion aberration DR distortion distortion aberration in dx x direction dz curvature of field δf Frequency interval (phase difference) δt aperture time difference of light intensity distribution δT Time interval at the light intensity center position measured at the aperture r0 central opening center position r1 opening center position W1-W4 Line width (reference pattern)

Claims (4)

[Claims] 1. A reference reticle having a plurality of reference patterns formed on an object plane of a projection optical system, a mask having a plurality of openings on an image plane of the projection optical system, and a light transmitting through the openings. A photodetector for measuring the amount of light to be emitted, and a moving mechanism for moving the mask in parallel and in a direction perpendicular to and parallel to the optical axis. In the method for measuring the amount of distortion aberration for measuring, the mask is reciprocally scanned in a direction perpendicular to and parallel to the optical axis to perform periodic motion, and the amount of light measured at an aperture provided on the optical axis of the projection optical system. Of the light quantity measured at each aperture provided outside the optical axis of the projection optical system is detected, and the distortion and the field curvature are measured by this phase difference. Distortion collection characterized by performing How to measure the difference. 2. The temporal distribution of the amount of received light at the openings provided on the optical axis of the projection optical system and the individual openings provided outside the optical axis of the projection optical system when the mask is moved cyclically. The phase difference from the temporal distribution of the amount of light measured at is used to develop the time distribution of the amount of received light at each aperture, and the frequency distribution of the amount of received light is used. The method for measuring the amount of distortion and the amount of curvature of field according to claim 1, wherein the amount of received light is calculated as the difference between the center intervals of the time distribution of the amount of received light. 3. The reference pattern provided on the reference reticle corresponding to the plurality of openings has a plurality of patterns of different shapes for the respective openings, and the reference patterns of the plurality of shapes. 3. The method for measuring the amount of distortion aberration and the amount of field curvature according to claim 1 or 2, wherein the distortion aberration and the field curvature of the projection optical system with respect to a predetermined reference pattern are measured. 4. A reference reticle having a plurality of reference patterns formed on the object plane of the projection optical system, a mask having a plurality of openings on the object plane of the projection optical system, and transmitting through the openings. A photodetector for measuring the amount of light to be formed, and a moving mechanism for moving the mask in a direction perpendicular to the optical axis,
In the method of measuring the amount of distortion of a projection optical system from the output of the time distribution of the light amount of the photodetector, the distortion in the state where the reference reticle and the mask are rotated and laterally displaced with respect to the projection optical system. From the measured value of the aberration and the measured value of the reference pattern interval and the aperture interval, the distortion aberration caused by the pattern arrangement error of the reference reticle and the mask is separated from the measured distortion aberration amount, and correction is performed. A method of measuring the amount of distortion.