JPH09251945A - Pattern for overlay accuracy control and overlay accuracy control method using the same - Google Patents

Abstract

(57) Abstract: In superposition accuracy control, a measurement parameter that can give a measurement value extremely close to a true misalignment amount is set in a short time. SOLUTION: A measurement is performed when a displacement of a known amount is set in advance in a pattern for overlay accuracy control, and the above displacement is measured most accurately when this pattern is measured using an overlay accuracy measuring device. Select a parameter and use it as the setpoint for subsequent measurements. Such a pattern is composed of two unit patterns 10 and 20 which are closely arranged with a minute distance d _10/20 to the extent that influences such as lens aberration and stage attitude error can be eliminated.
There is no misalignment between the lower layer side pattern 1a and the upper layer side pattern 2a, but the unit pattern 20 sets a known misalignment amount S ₁ . A measurement parameter is selected when the difference between the measurement values of the misalignment amounts of both unit patterns 10 and 20 is closest to S ₁ .

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a pattern and a method for overlay accuracy control capable of detecting an absolute error in the overlay accuracy control of upper and lower patterns formed by photolithography.

[0002]

2. Description of the Related Art Lithography is a technique for increasing the NA of the projection lens (objective lens) of a reduction projection exposure apparatus, shortening the exposure wavelength seen in the transition from the bright line of a high-pressure mercury lamp to excimer laser light, or phase shifting. The resolution has been steadily improved through the adoption of super-resolution technology such as the method and modified illumination method.

On the other hand, in response to this, the technique for managing the overlay accuracy is also advancing. However, the improvement in the overlay accuracy up to now is the result of steady improvement of the apparatus performance such as the damping and temperature control of the exposure stage and the aberration of the projection optical system. Generally, the overlay accuracy is 1 / the design rule (minimum processing size).
A value of 3-1 / 4 is required, and after the 0.3 μm rule, a high overlay accuracy of 0.1 μm or less is required as a management target value. However, it is becoming increasingly difficult to achieve such high precision only with the conventional improvement policy, and it is said that the limitation of the exposure technique lies in the overlay technique.

Generally, the control items of overlay accuracy include measurement reproducibility 3σ and TIS (Tool Induced Shift: shift caused by the measuring device).

The measurement reproducibility is determined by repeating the measurement of the overlay accuracy for a plurality of overlay accuracy control patterns formed on the substrate a plurality of times, and obtaining a variation in measured values for the same pattern (ie, standard deviation σ). Is the evaluation standard. It can be said that the larger the value of 3σ, the lower the accuracy of the overlay accuracy measuring device.

On the other hand, in the TIS, after measuring the position coordinates (X ₀ , Y ₀ ) of a certain overlay accuracy control pattern, the substrate is rotated 180 ° and the position coordinates (X X) of the same pattern are measured.
_180, Y _{1 80)} the average position coordinates when measuring ((X ₀ + X
₁₈₀ ) / 2, (Y ₀ + Y ₁₈₀ ) / 2).
If X ₀ = −X ₁₈₀ and Y ₀ = −Y ₁₈₀ ideally hold, the average position coordinate should be (0,0),
Actually, the imperfections of the optical system of the overlay accuracy measuring device,
In particular, it is usually a value that has a small absolute value that reflects the insufficient verticality of the optical axis of the illumination light, and the larger this absolute value, the lower the accuracy of the overlay accuracy measurement device. .

As an example of the above-mentioned overlay accuracy control pattern, a so-called box-in-box (B
ox-in-Box) type pattern. This pattern has a lower layer side pattern 12 which is a rectangular (here, square) concave portion formed on the substrate 11, as shown in a top view (a) and a cross-sectional view taken along line CC of FIG. , A combination with the upper layer side pattern 14 which is a rectangular (also square) convex portion formed inside the concave portion. Here, it is assumed that the upper layer side pattern 14 is a photoresist pattern. Therefore, the base of the upper layer side pattern 14 is deposited on the entire surface of the substrate 11 and processed by using the upper layer side pattern 14 as a mask. It is a film to be processed 13sym (subscript sym represents that the film to be processed symmetrically covers the lower layer side pattern 12).

Measurement of such a box-in-box type pattern is generally performed by the following method. That is, with the optical axis of the illumination system of the overlay accuracy measuring device aligned with the center of the pattern, the pattern is irradiated with light having a predetermined wavelength and intensity, and the reflected light is captured by a linear CCD and converted into an electrical signal. Then, the electric signal waveform as shown in FIG. 3C is obtained, the midpoint positions between the edges of the upper and lower patterns are respectively determined from this waveform based on the threshold method, and the two are compared.

This method will be specifically described with reference to FIG. 1B. The distance between edges of the upper layer side pattern 14 is d ₁₄ , the center position between the edges is C ₁₄ , and the lower layer side pattern 1 is described.
The distance between the two edges is d ₁₂ , and the center position between the edges is C _12.
In this case, if the upper and lower patterns have no misalignment, the center positions C ₁₂ and C ₁₄ between the edges match each other, but if the misalignment occurs as shown in the figure, the two do not match. Misalignment amount St at this time
With respect to (the subscript t represents a true value), 3σ and TIS are examined.

By the way, the misalignment amount St measured as described above is not an absolute error, that is, a true misalignment amount, no matter how good the values of 3σ and TIS are. This is because, in addition to the true misalignment amount actually occurring on the substrate, the measurement error inherent in the overlay accuracy measuring device is inevitably added to the measured misalignment amount. . Therefore, it is highly possible that the wrong value is read with good reproducibility.

Under the conventional design rule of the half-micron level or the sub-half-micron level, it is not always necessary to know the true misalignment amount. Therefore, conventionally, great attention should be paid to the true misalignment amount. Was never paid. However, in the future quarter-micron level or finer levels thereafter, the ratio of the overlay deviation amount to the minimum resolution dimension increases, so the measurement error inherent in the overlay accuracy measurement device is eliminated as much as possible, and the measured value It is necessary to make efforts to bring the value closer to the true value.

In order to deal with the above problem, the applicant of the present application has previously disclosed in Japanese Patent Laid-Open No. 6-132189 that the measurement conditions of the overlay accuracy measuring device are optimized without measuring the true overlay deviation amount. Suggesting a way to do it.
In this method, first, the misalignment amount (ΔXs, ΔYs) generated in the peripheral measurement marks arranged in the peripheral part within one shot of photolithography and the misalignment amount (ΔXs, ΔYs) generated in the central measurement mark arranged in the central part ( ΔXc, ΔYc) and measuring the respective displacement superimposed by subtracting the latter from the former amount difference (ΔXs-ΔXc, ΔYs-ΔYc ) is calculated, obtaining the variation of the difference (3σx _a, 3σy _a). By the above operation, the effects of the aberration of the projection lens of the exposure apparatus, the thermal expansion and contraction of the substrate, and the rotation (rotation) of the substrate are canceled out, and eventually, the variations (3σx _a , 3σy _a ) in the misalignment amount of the overlay are superposed with each other. This corresponds to the overlay measurement error inherent in the measuring device. The measurement value can be brought close to the true value by adjusting the measurement condition of the overlay accuracy measuring device so that this variation is minimized.

[0013]

By the way, the method described in Japanese Patent Laid-Open No. 6-132189 mentioned above is premised on that "the amount of misalignment within a shot remains unchanged within the substrate surface". However, in reality, since the stage that supports the substrate causes posture errors such as yaw, pitch, and roll, strictly speaking, the in-shot component varies even within the substrate surface. In the above method, by measuring a large number of shots on the substrate and performing an averaging operation, the influence of this variation is reduced, but such a method is complicated,
The time required for measurement also becomes long. Further, the above method is, of course, sufficient from the viewpoint of the overlay accuracy measurement condition, that is, the optimization of the parameter setting of the overlay accuracy measurement device, but it is still not possible to know the true value of the overlay deviation amount itself. .

Further, due to the asymmetry of the underlayer film on the substrate, the accuracy of measurement by the overlay accuracy measuring device may be limited. This problem will be described with reference to FIG. This figure is box-in
The box-shaped pattern is shown as an example, in which (a) is a top view, (b) is a sectional view taken along the line CC, and (c) is obtained when this pattern is measured by an overlay accuracy measuring device. It is an electric signal waveform that is generated. The difference between this figure and FIG. 3 is that the processed film 13asym covering the lower layer side pattern 12 is formed.
(The subscript asym indicates that it is asymmetric.)
The coverage is asymmetric. This asymmetry is caused, for example, by the influence of a divergent magnetic field in the vicinity of the substrate in the ECR-CVD method, or by the influence of the deviation of the incident direction of the deposited species on the substrate in the sputtering method.

The true amount of misalignment St occurring in this pattern is determined by the center position C ₁₄ between edges corresponding to the midpoint of the distance d ₁₄ between edges of the upper layer side pattern 14 and the distance between the edges of the lower layer side pattern 12. It is defined as the distance to the center position C ₁₂ between edges corresponding to the midpoint of d ₁₂ . However, when the film thickness of the film to be processed 13asym is different at both edges of the lower layer side pattern 12, the edge distance d _12a of the lower layer side pattern 12 detected by the threshold method is, for example, the true edge distance d as shown in the figure. _If it is smaller than _{12, the} midpoint position C _12a between the edges shifts accordingly. As a result, the measured misalignment amount Sm (subscript m represents a measured value), that is, the center position C ₁₄ between the edges of the upper layer side pattern 14 and the middle position C _12a between the edges of the lower layer side pattern 14 is measured. The distance between them becomes smaller than the true value St. In this case, no matter how the measurement parameters are optimized, the measurement value naturally has a limit in approaching the true value.

Under such circumstances, in reality, in order to know the true value of the misalignment amount, it is necessary to rely on another means such as a length measuring SEM (scanning electron microscope) or device characteristic measurement. However, it is difficult to apply these means to a substrate in the process at any time. Therefore, the present invention provides a pattern for overlay accuracy control that can set a measurement parameter that can give a measured value extremely close to the true misalignment amount in a short time without using these separate means, and an overlay accuracy management using the pattern. The purpose is to provide a method.

[0017]

According to the present invention, when a pattern for superposition accuracy control in which a known amount of displacement is preliminarily formed is formed on a substrate and the pattern is actually measured by a superposition device, the displacement is measured. It is intended to achieve the above-mentioned object by setting the measurement parameters that can measure the most accurately and performing the subsequent measurement.

Here, the above-mentioned known amount of displacement may be set between a plurality of unit patterns arranged in close proximity, or may be set inside a single pattern. In other words, when a displacement is set between a plurality of unit patterns, in other words, a plurality of unit patterns each including a combination of a lower layer side pattern and an upper layer side pattern having a geometrical relationship with the lower layer side pattern are arranged close to each other. In this case, the relative position of the lower layer side pattern and the upper layer side pattern in at least one of the unit patterns is displaced from the normal position by a known amount. On the other hand, when the displacement is set inside the single pattern, one of the lower layer side pattern and the upper layer side pattern has an edge distance that changes in at least two steps, and at least the midpoint position between the edges is set. This is a case where one is displaced from the midpoint position between the edges of the other pattern by a known amount.

[0019]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below.

First Embodiment Here, in the box-in
An overlay accuracy management pattern in which a box-shaped pattern and a box-in-box pattern having a known misalignment amount are closely arranged, and an overlay accuracy management method using the same will be described with reference to FIG. .

This figure shows a superposition accuracy control pattern in which two unit patterns 10 and 20 each having a square shape are adjacently arranged at a distance d _10/20 . The inter-edge distances d _1a and d _1b of the lower layer patterns 1a and 1b of both unit patterns 10 and 20 are both 10 to 30 μm and equal to each other.
Further, the distances d _2a and d between the edges of the upper layer patterns 2a and 2b
Both _2b are 5 to 20 μm and are equal to each other. Here, in the pattern 10 on the left side in the drawing, the midpoint position C _1a between edges of the lower layer side pattern 1a and the midpoint position C _2a between edges of the upper layer side pattern 2a coincide. That is, there is no design deviation in the photomask (normal position). On the other hand, in the pattern 20 on the right side, a known misalignment amount S ₁ is set between the midpoint position C _1b between edges of the lower layer side pattern 2a and the midpoint position C _2b between edges of the upper layer side pattern 2b. There is. That is, it is a state in which the displacement is generated from the normal position. This misalignment amount S _{1 is set} to 0.05 μ
m. That is, the design deviation in the photomask is 0.25 μm when the reduction ratio of the stepper is 1/5. The distance d _10/20 is 20 to 30 μm, which is a minute distance sufficient to eliminate the effects of aberration of the reduction projection lens of the stepper, attitude error of the wafer stage, measurement error when checking the reticle accuracy, and the like. ing.

In the measurement using the overlay accuracy control pattern as described above, the unit patterns 10, 2 are moved while the stage is moved using a conventional overlay accuracy measuring device.
It is carried out sequentially by detecting the center of 0 and irradiating with measurement light. At this time, if the measurement is ideally performed,
(Measured value of misalignment amount of unit pattern 20)-(Measured value of misalignment amount of unit pattern 10) = 0.05 μm
It should be. Therefore, if the measurement parameter is found when the measurement value closest to this value is obtained, and the subsequent measurement is performed by setting this measurement parameter in the above device,
It becomes possible to obtain a measured value that is substantially equal to or very close to the true overlay deviation amount. This measurement parameter is at least one of the focus position, the numerical aperture of the lens, the wavelength of illumination light, the amount of illumination light, and the optical axis position.

As described above, according to the present invention, the measurement parameter is selected in the case where the displacement of the known amount preset in the overlay accuracy control pattern is correctly measured. Therefore, even if only the unit pattern 20 is used alone for the time being, The measurement parameters can be optimized. However, in the case of such single use, it is necessary to take measures such as inputting an offset value in advance in a related file of software of the overlay accuracy measuring device, which is complicated. 2 as described above
The reason why one unit pattern is used is to eliminate such complication and enable the existing device to be used as it is.

The number of unit patterns arranged in proximity is not particularly limited, and the distance between them is such that the aberration of the reduction projection lens of the stepper, the attitude error of the wafer stage, and the measurement error when confirming the reticle accuracy. As long as they are not affected by the above, they may have any positional relationship with each other. However, as the number of unit patterns increases, the area occupied by the overlay accuracy control pattern also increases, and since the stage must be moved to the location of each pattern and the same number of measurements must be performed,
In consideration of this, it is necessary to determine the practically preferable number. In principle, the present invention can be implemented with two.

Although a known amount of displacement may be set for each of the plurality of unit patterns, at least one of the unit patterns has a lower layer side pattern and an upper layer side pattern as in the above-described example. If the relative position is a positive position, that is, the misalignment amount is set to zero, the misalignment amounts measured for other unit patterns directly represent the true misalignment amount, which is convenient.

Although the upper layer side pattern is typically a photoresist pattern, it may be a pattern of a film to be processed which is etched by using the photoresist pattern as a mask. However, in the latter case, of course, the overlay accuracy including the etching accuracy will be controlled, and the overlay misalignment that occurs during the photolithography stage will be found before etching to regenerate the resist pattern. Such management cannot be done.

Second Embodiment In the present embodiment , a pattern for overlay accuracy control that allows a known amount of displacement to be included in the pattern by devising the pattern shape itself and enables detection of the true misalignment amount even when used alone, and this pattern A superposition accuracy control method using is described with reference to FIG.

As is apparent from FIG. 2, this pattern is a modification of the ordinary square box-in-box pattern. The idea is that by overlaying two upper layer side patterns that are in a positional relationship with each other as shown in FIG. Is to let. That is, although the shape of the lower layer side pattern 3 is a square, the upper layer side pattern 4 to be superposed on this is
Is substantially L-shaped, that is, the distance between edges changes in two steps along the horizontal direction in the figure. This upper layer side pattern 4 has a region 4a having a distance d _4a between edges,
It is composed of a region 4b having a distance d _4b between edges, and a known misalignment amount S ₂ is set between the midpoint position C _4a between edges of the region _4a and the midpoint position C _4b between edges of the region 4b. . Further, of the two types of midpoint positions C _4a and C _4b between edges of the upper layer side pattern 4a, the latter midpoint position C between edges C
_4b coincides with the midpoint position C ₃ between the edges of the lower layer side pattern 3.

The measurement using the overlay accuracy control pattern as described above is performed after changing the software of the conventional overlay accuracy measuring device. This cannot be performed by the conventional method of detecting the pattern center and irradiating the pattern with measurement light, and the pattern is automatically measured in a certain direction (vertical direction in FIG. 2) with respect to the detected pattern center. This is because it is necessary to apply a constant amount of offset and irradiate each offset point with the measurement light. When applying this offset, it is most effective to eliminate the influence of aberration by actually moving the stage to keep the optical axis of the measurement optical system always perpendicular to the pattern, but do not move the stage. It is also possible to simultaneously capture signals from both points. As a result, a signal waveform corresponding to the cross section A and a signal waveform corresponding to the cross section B in the figure are obtained. From these two signal waveforms, the midpoint positions C _4a and C between the edges
_4b is detected, a measurement parameter when the misalignment amount corresponding to the distance between the _two is the closest to the known misalignment amount S ₂ is found, and this measurement parameter is set in the above device to perform subsequent measurements. By doing so, it becomes possible to obtain a measured value that is extremely close to the true overlay deviation amount. This measurement parameter is at least one of the focus position, the numerical aperture of the lens, the wavelength of illumination light, the amount of illumination light, and the optical axis position.

The overlay accuracy control pattern according to the present embodiment requires a much smaller stage movement amount than the pattern shown in FIG. Further, since the patterns can be used independently, the occupied area can be reduced as compared with the case of forming a plurality of unit patterns.

In the above pattern, both the two midpoint positions C _4a and C _4b between the edges of the upper layer side pattern 4a should be displaced from the midpoint position C ₃ between the edges of the lower layer side pattern 3. You can also However, if the midpoint position C _4b between the edges is made to coincide with the midpoint position C ₃ between the edges as described above, the misalignment of the upper and lower patterns can be known very easily.

Although the present invention has been described above based on the two embodiments, the present invention is not limited to these embodiments. For example, if the overlay accuracy control pattern of the present invention has some geometrical relationship between the upper layer side pattern and the lower layer side pattern, and that a known amount of displacement can be set inside and outside the pattern, its shape It doesn't matter. For example, the lower layer side pattern of the box-in-box type pattern may be formed by a convex portion instead of the concave portion as described above. Alternatively, it may be a pattern in which a lower layer side pattern and an upper layer side pattern are arranged in parallel, instead of a pattern of a nested structure such as a box-in-box type.

[0033]

As is apparent from the above description, according to the present invention, it is possible to set the optimum measurement parameter in overlay accuracy control in a short time, and to obtain a value very close to the true misalignment amount. Can be obtained. The present invention improves the reliability of semiconductor devices and liquid crystal devices manufactured based on minute design rules, reduces costs,
This greatly contributes to the improvement of throughput.

[Brief description of drawings]

FIG. 1 is a top view of a superposition accuracy control pattern of the present invention in which a known amount of displacement is set between two unit patterns arranged in proximity.

FIG. 2 is a top view of an overlay accuracy control pattern of the present invention in which a known amount of displacement is set inside a single pattern.

3A and 3B are diagrams illustrating an ideal example of a conventional misalignment amount detection method using a box-in-box type pattern. FIG. 3A is a top view of the pattern, and FIG.
The -C line sectional view and (c) diagram respectively show the electric signal waveforms obtained by the measurement.

4A and 4B are views for explaining a problem in detecting a misalignment amount using a conventional box-in-box type pattern, FIG. 4A is a top view of the pattern, and FIG. 4B is its CC. The line sectional view and (c) diagram respectively show the electric signal waveforms obtained by the measurement.

[Explanation of symbols]

10, 20 unit patterns 1a, 1b, 3 lower layer side patterns 2a, 2b, 4 upper layer side patterns S ₁ , S ₂
Known misalignment amount

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁶ Identification code Agency reference number FI Technical display location H01L 21/30 525W

Claims (12)

[Claims] 1. An overlay accuracy control pattern in which the relative position of the lower layer side pattern and the upper layer side pattern having a geometrical relationship with the lower layer side pattern is displaced from the normal position by a known amount. 2. A plurality of unit patterns composed of a combination of a lower layer side pattern and an upper layer side pattern having a geometrical relationship with the lower layer side pattern are arranged in close proximity, and at least one of the unit patterns is the lower layer. An overlay accuracy control pattern in which the relative position between the side pattern and the upper layer side pattern is displaced from the normal position by a known amount. 3. The overlay accuracy management pattern according to claim 2, wherein a relative position between the lower layer side pattern and the upper layer side pattern is a positive position in at least one of the unit patterns. 4. The overlay accuracy management pattern according to claim 2, wherein the plurality of unit patterns are arranged close to each other with a minute distance therebetween. 5. The normal position is defined by a coincidence between a middle point position between edges of the lower layer side pattern along a certain direction and a middle point position between edges of the upper layer side pattern along the same direction, and the displacement is defined. 3. The overlay accuracy control pattern according to claim 2, wherein is defined by the disagreement of the midpoint positions between these edges. 6. The upper layer side pattern is a photoresist.
The overlay accuracy management pattern according to claim 2, which is a pattern. 7. A combination of a lower layer side pattern and an upper layer side pattern having a geometrical relationship with the lower layer side pattern, wherein either one of these patterns has an interval between edges which changes in at least two steps along a certain direction. Have a distance,
And, at least one of the midpoint positions between the edges is a superposition accuracy management pattern in which the midpoint position between the edges of the other pattern is displaced by a known amount. 8. One of the midpoint positions between the edges of the lower-layer side pattern and the upper-layer side pattern, which has an inter-edge distance that changes in at least two steps along a certain direction, is located between the edges of the other pattern. The overlay accuracy management pattern according to claim 7, wherein the pattern is coincident with the midpoint position. 9. The overlay accuracy between the upper and lower patterns on the substrate is measured by measuring at least one overlay accuracy control pattern on the substrate for each photolithography shot by using an overlay accuracy measuring device. Is a method for managing overlay accuracy, wherein a displacement of a known amount is set in advance in the overlay accuracy management pattern, and a measurement parameter capable of measuring this known amount most accurately is set in the overlay accuracy measuring device. A method of controlling the overlay accuracy for performing measurements. 10. The overlay accuracy control pattern is composed of a plurality of unit patterns which are composed of a combination of a lower layer side pattern and an upper layer side pattern having a geometrical relationship with the lower layer side pattern and which are closely arranged on the substrate. 10. The overlay accuracy management method according to claim 9, wherein the known amount of displacement is set between the unit patterns. 11. The overlay accuracy management method according to claim 9, wherein the known amount of displacement is set inside a single overlay accuracy management pattern. 12. The overlay accuracy management method according to claim 9, wherein the measurement parameter is at least one of a focal position, a numerical aperture of a lens, a wavelength of illumination light, an illumination light amount, and an optical axis position.