JPH09199414A - Methods for measuring alignment error and expositing and for controlling overlay accuracy in semiconductor device manufacturing process - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the methods for measuring an alignment error and exposing and for controlling an overlay accuracy which can increase a throughput, keeping a processing accuracy and a controlling accuracy as they are. SOLUTION: Since an alignment error between fields and an in-field alignment error can be handled separately in theory and a variation in in-field component error such as a magnification error and rotation error is usually small enough to be neglected, there is no necessity of multi-point measurement of all the exposure fields sampled from a wafer W as in the conventional method and a multi-point measurement of only a few exposure fields is good enough to realize a practical accuracy. Nine exposure fields F are sampled from one wafer W and then a five-point in-field measurement is conducted for a central field and a one-point in-field measurement is conducted for eight fields in the periphery of the wafer W. The total number of measurement points P is 13 and this number is remarkably fewer than 45 in the conventional method.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to measurement of alignment accuracy between a substrate and a photomask in a manufacturing process of a semiconductor device formed on a substrate by using a lithography technique, exposure, or overlay between upper and lower patterns. The present invention relates to a technique for improving throughput in management of alignment accuracy.

[0002]

2. Description of the Related Art Lithography applied to microfabrication such as semiconductor device manufacturing is concerned with the high NA of a projection lens (objective lens) of a reduction projection exposure apparatus and the shift from the bright line of a high pressure mercury lamp to an excimer laser beam. The resolution has been steadily improved by shortening the exposure wavelength used, or by adopting a super-resolution technique such as a phase shift method or a modified illumination method.

On the other hand, also in the technique of alignment accuracy, alignment accuracy has been improved by improving the apparatus performance such as vibration control and temperature control of the exposure stage or improvement of aberration of the projection optical system. As for the alignment accuracy,
A value of 1/3 to 1/4 of the rule (minimum processing dimension) is required, and for example, in the 0.3 μm rule, a high alignment accuracy of 0.1 μm or less is required as a management target value. And the technology that enables such highly accurate alignment operation is
In addition, there is a demand for a control method of overlay accuracy.

First, in the exposure process, alignment is performed based on the statistical processing of the deviation between the stage coordinate system and the substrate coordinate system measured by sampling shots. For example, by introducing a conversion model based on a linear equation, the field to be exposed on the substrate (hereinafter, referred to as an exposure field or simply field) on the substrate prior to exposure is
From the simultaneous equations obtained by partially differentiating the error component described by the sum of squared differences of the measured coordinates and the reference coordinates with the respective conversion parameters of the first-order conversion model, the respective conversions of the first-order conversion model Exposure is performed after parameters are defined, an alignment grid (predictive grid) is created based on a first-order conversion model having the conversion parameters defined as coefficients, and alignment is performed based on the coordinates indicated by the alignment grid.

Further, the general superposition accuracy management steps are as follows. (1) A field in which a predetermined number of wafers are extracted from one lot and several exposures have been performed in each wafer (hereinafter, this exposed field is also referred to as an exposure field for convenience. Is described each time it is necessary to distinguish)
The overlay error occurring in the sampled field is measured by visual observation of vernier or by using a dedicated measuring device, and (2) the average value x and the standard deviation σ are obtained from the overlay error. (3) | x | +3
σ is compared with the management target value (spec), and (4) | x | +
If it is the 3σ management target value, it is determined to be passed, and if | x | + 3σ> the management target value, it is determined to be unacceptable (spec out). As described above, in the control of the overlay accuracy, the value of | x | + 3σ (x is the average value, | x | is the absolute value of the average value, and σ is the standard deviation) is used as the management target value, as in a general quality control method. It is done by comparing with.

Here, | x | + 3σ is a range including 99.7% of the population in the normal distribution. | X | + 3σ
If the control target value is less than the control target value, all the wafers in the lot are sent to the next process. If the control target value is exceeded (spec out), all the wafers in the lot are sent to the remanufacturing process. It Here, the regenerating step refers to a series of steps in which the resist pattern is once peeled off and lithography is performed again to re-form the resist pattern.

The technique commonly required for the exposure process and the overlay accuracy control process as described above is a technique for measuring an alignment error. Hereinafter, the error relating to the alignment in the exposure process or the overlay error in the overlay accuracy control process will be collectively referred to as “alignment error” for convenience of description, unless otherwise specified. The same applies to "alignment accuracy". By the way, in the past, as such an alignment error, an alignment error mainly occurring between fields has been targeted. For example, regarding the management of the alignment accuracy, the error between fields was the target. However, in recent years, semiconductor devices are highly integrated, and a plurality of chips are exposed in one shot to improve productivity. Therefore, the size of each exposure field is 20 mm.
The size is larger than a corner, and as a result, the alignment error at the corner of the exposure field reaches a level that cannot be ignored in both the exposure process and the overlay accuracy control process.

Therefore, in addition to the field alignment accuracy, a technique related to the field alignment accuracy is
It is important in the exposure process and the overlay accuracy control process. In other words, “total overlay accuracy”, which means a total alignment accuracy that combines the alignment accuracy inside and outside the field, has been required.

By the way, the measurement of the inter-field alignment accuracy as described above has usually been carried out by so-called one-field measurement in which one point is selected in one exposure field and the alignment error is measured. . On the other hand, in-field alignment accuracy is usually measured by so-called multi-field measurement, in which a plurality of points are selected in one exposure field to measure alignment error.

[0010]

By the way, when the above-mentioned multi-point measurement in the field is performed by using the existing exposure apparatus or overlay accuracy measuring machine, the exposure field sampled for measuring the inter-field accuracy is used. Multipoint measurement is performed for each of the. In this case, the exposure apparatus targets all the wafers that make up the lot, and the overlay accuracy measuring apparatus targets the wafers sampled from the lot. For example, FIG. 5 shows processing by such an overlay accuracy measuring machine. In the figure, for each of, for example, three wafers W sampled from one lot, five measurements are made for each of all nine exposure fields F sampled from a total of 25 exposure fields F. The manner in which the measurement is performed at the point P is schematically shown.

That is, the total number of measurement points is (5
It is 45 in points) × (9 fields). Therefore, if five-point measurement is performed on all the exposure fields F sampled as described above, the time required for measurement is five times or more as compared with the case where only the inter-field overlay error is measured by one-point measurement. There was a problem that it would increase. This is because it takes time to handle and align the wafer in addition to the time required for the measurement itself. The above-mentioned problems similarly occur in the measurement of the in-field component alignment in the exposure process.

As described above, the conventional technique has a problem that the throughput of lithography is significantly reduced due to the increase of sampling points, resulting in an increase in the cost of semiconductor devices. Therefore, the present invention solves the above-mentioned problems of the prior art, and is capable of improving throughput while maintaining processing accuracy and management accuracy, and a method for measuring alignment accuracy and an exposure method,
It is an object to provide a superposition accuracy control method.

[0013]

A method of measuring alignment accuracy according to the present invention is proposed to achieve the above-mentioned object, and is a conventional method of performing multipoint measurement on all sampled exposure fields. The number of exposure fields for multi-point measurement (hereinafter referred to as multi-point measurement fields) is reduced by at least 1 from the total number of sampled exposure fields, and one-point measurement is performed for the remaining exposure fields. That is the basic idea. This is based on the theoretical basis that "the intra-field alignment error and the inter-field alignment error can be treated independently" and the knowledge that the present inventor has obtained empirically based on actual measurement using a mechanical sample. To do.

First, the above-mentioned theoretical basis will be described by taking an overlay error as an example. Now, consider the measurement of overlay error in the exposure field as shown in FIG. This figure shows the relative error between the exposure field serving as the measurement reference and the exposure field in which the overlay error has occurred, and is an example of 5-point measurement in the field. Where i is the reference
The coordinates (x
ij, yij) and the coordinates (xij ', jj of the measurement point in the i-th exposure field where the overlay error has occurred)
yij ′), the following relation of the following equation 1 is established.

[0015]

[Equation 1]

In Equation 1, Tx and Ty represent parallel movement, M represents magnification, and θ represents rotation. All of these are linear error components. The linear error is a correctable first-order error that appears on the wafer W with a certain tendency. When discussing the overlay error between fields, the feature is that the scaling factor (the scaling factor between fields is called “scaling”) and the rotation increase in the radial direction from the center of the wafer W. When discussing the overlay error in the field as described above, it is treated as a component that does not change independently in the xy directions.

When the variance σ2 (the standard deviation σ squared) is calculated from the equation 1, the equation 2 is obtained.

[0018]

[Equation 2]

However, in the above discussion, the non-linear error in the field is not taken into consideration. The non-linear error is a second or higher order error on the wafer W, and typically includes an error due to the aberration of the projection lens of the exposure apparatus, the thermal deformation of the wafer W, and a measurement error at the time of overlay error measurement. Such a non-linear error uses a plurality of exposure apparatuses having different resolutions according to the design rule of each layer, not to say so-called mix & match exposure, but uses a single exposure apparatus for all layers, Even when so-called single machine exposure is performed, a considerable amount occurs.
Therefore, as a more realistic expression of total overlay, Equation 3 is appropriate.

[0020]

(Equation 3)

Here, the term relating to the intra-field superposition is summarized and the following equation 4

[0022]

(Equation 4)

In summary, the equation 3 can be rewritten as the following equation 5.

[0024]

(Equation 5)

Equation 5 is the total overlay variance σt
total2 is the variance σi of the overlay error in the field
ntra2 and inter-field overlay error variance σin
It is represented by the sum of ter2. That is,
It can be seen that overlay errors inside and outside the exposure field can be treated theoretically independently of each other.

Next, the empirical knowledge obtained by the inventor from the actual measurement of the mechanical sample will be described. What is a mechanical sample?
A typical process included in an actual semiconductor device manufacturing process is reproduced on a wafer. The study by the present inventor has confirmed that variations in intra-field errors such as magnification and rotation are so small that they can be ignored. As a result, in Expression 4, the term relating to dispersion can be ignored, and the magnification and rotation can be represented by the magnification and rotation in the i-th exposure field. Based on the above knowledge, Equation 4 can be rewritten as Equation 6 below.

[0027]

(Equation 6)

In other words, it is not necessary to carry out multipoint measurement for all of the sampled exposure fields, and multipoint measurement may be carried out only for some exposure fields and one point measurement may be carried out for the remaining exposure fields. ,
It shows that correct evaluation is possible.

Based on the above theory and knowledge, and in order to achieve the above object, the alignment error measuring method according to the present invention provides a plurality of substrates on a substrate corresponding to repeated lithographic shots on the substrate in the semiconductor device manufacturing process. In a method of measuring an alignment error generated for an exposure field to be formed, m (m is 1 ≦ 1) of n (n is a natural number of 2 or more) exposure fields sampled from the substrate.
It is characterized in that in-field multipoint measurement is performed for natural numbers satisfying m ≦ n−1, and in-field one-point measurement is performed for the remaining (n−m) numbers. In addition, the above-mentioned m is set to 1, in particular.

According to the above-described alignment error measuring method of the present invention, the number of multi-point measurement exposure fields is reduced in substrate (wafer) units.

The exposure method according to the present invention is constructed according to the alignment error measuring method. That is, the exposure method according to the present invention is an exposure method in which alignment is performed based on measurement of a predetermined portion of an exposure field formed in plural on the substrate corresponding to repeated lithography shots on the substrate in the semiconductor device manufacturing process. In, n samples from the substrate (n is a natural number of 2 or more)
In the exposure field of, the multi-point measurement in the field is performed for m (m is a natural number satisfying 1 ≦ m ≦ n−1), and the single point measurement in the field is performed for the remaining (n−m), and In-field component alignment of at least one exposure field is performed based on the multi-field measurement result. In addition, especially m
Is set to 1.

According to the above-described exposure method of the present invention, in-field linear error correction is simultaneously performed for each of the other exposure fields by only performing alignment correction for one shot or several shots within the same wafer. By doing so, the number of exposure fields for multipoint measurement is the substrate (wafer).
Decrease in units.

Further, according to another empirical finding obtained by the present inventor from actual measurement of another mechanical sample, the difference in the in-field linear error between wafers belonging to the same lot is negligibly small. In other words, it was confirmed that the field magnification and the field rotation, which are the linear error components in the field, were so small that the fluctuation amount in the same lot was negligible in the measured values. Therefore, according to this result, for a plurality of wafers belonging to the same lot, the wafer 1
It was clarified that the error measurement of the in-field component was performed on only one wafer or every few wafers, and that the measurement result can be applied to other wafers belonging to the same lot.

Based on the above findings, the alignment error measuring method according to the present invention provides the alignment error that occurs in the exposure field formed in plural on the substrate in response to repeated lithography shots on the substrate in the semiconductor device manufacturing process. In the measurement method of (1), M (M is 1 ≦ 1) of N (N is a natural number of 2 or more) substrates that compose the lot to which the substrate belongs.
For a substrate of a natural number satisfying M ≦ N−1, multipoint measurement within a field is performed on at least one of a plurality of exposure fields sampled from each substrate,
For the remaining (N−M) substrates, one point in-field measurement is performed for a plurality of exposure fields sampled from each substrate. In addition, it is characterized in that multi-field measurement is performed for all of the plurality of exposure fields sampled for the M substrates.

According to the above-described alignment error measuring method according to the present invention, the number of multi-point measurement exposure fields is reduced in lot units.

In the exposure method according to the present invention, which is constructed according to the above-mentioned alignment error measuring method, a plurality of exposure methods are formed on a substrate corresponding to repeated lithographic shots on the substrate in the semiconductor device manufacturing process. In the exposure method in which the alignment is performed based on the measurement of a predetermined portion of the exposure field, all N (N
Is a natural number greater than or equal to 2), a plurality of M samples (M is a natural number satisfying 1 ≦ M ≦ N−1) including the first substrate of the lot are sampled on each of the substrates. In-field multi-point measurement is performed for at least one of the exposure fields, and in-field multi-point measurement is performed for the plurality of exposure fields sampled from each of the remaining (NM) substrates. On the basis of the in-field multipoint measurement results of the M substrates, in-field component alignment is performed on the exposure field of the substrates, and the in-field multipoint measurement results of the M substrates are It is also characterized in that it is applied to (N−M) substrates to perform in-field component alignment of an exposure field. In addition, it is characterized in that multi-field measurement is performed for all of the plurality of exposure fields sampled for the M substrates.

According to the above-described exposure method of the present invention, the number of exposure fields for multipoint measurement is reduced in lot units.

As described above, in the exposure method according to the present invention,
The alignment of the in-field component with respect to each substrate belonging to the same lot is performed by using the alignment correction amount of the in-field component already performed using any substrate including at least the first substrate belonging to the lot to other substrates. It is executed by applying the above method, but in this case, it is particularly preferable to apply the alignment correction amount obtained for the first substrate of the lot.

Next, in the overlay accuracy control method according to the present invention, overlay errors that occur in a plurality of exposure fields formed on the substrate are measured in response to repeated lithography shots on the substrate, and the measurement is performed. In the method of managing the overlay accuracy based on the result, in the overlay error measurement, m out of n (n is a natural number of 2 or more) exposure fields sampled from the substrate.
In-field multipoint measurement is performed for the number (m is a natural number satisfying 1 ≦ m ≦ n−1), and in-field one-point measurement is performed for the remaining (n−m). In addition, the above-mentioned m is set to 1, in particular.

According to the overlay accuracy management method according to the present invention, the number of exposure fields for multipoint measurement is reduced on a substrate (wafer) basis.

Alternatively, in the overlay accuracy control method according to the present invention, overlay errors occurring in a plurality of exposure fields formed on the substrate are measured in response to repeated lithography shots on the substrate, and the measurement is performed. In the method of managing the overlay accuracy based on the result, in the measurement of the overlay error, M (M is 1 ≦ M) of N (N is a natural number of 2 or more) substrates sampled from the lot to which the substrate belongs. For a number of substrates satisfying M ≦ N−1, multi-point measurement within a field is performed on at least one of a plurality of exposure fields sampled from each substrate, and for the remaining (N−M) substrates, It is characterized in that one field measurement is performed for a plurality of exposure fields sampled from each substrate. In addition, it is characterized in that multipoint measurement is performed for all of the plurality of exposure fields sampled for the M substrates. Or the above M
Is set to 1.

According to the overlay accuracy control method according to the present invention, the number of exposure fields for multipoint measurement is reduced in lot units.

[0043]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below based on examples. Example 1 Table 1 shows measured values of linear errors of in-field components (field magnification and field rotation) in each of exposure fields arranged in a cross direction on one wafer surface. The exposure field is centered on shot number 4, with the top exposure field of shot number 1 to the bottom exposure field of shot number 7 arranged vertically on the surface, and the left end exposure of shot number 8 centered on shot number 4. The fields from the field to the rightmost exposure field of shot number 13 are arrayed in the left-right direction on the surface, and the exposure is the result of performing alignment correction of the in-field components for each field.

[0044]

[Table 1]

As shown by the actual measurement values of the field magnification and the field rotation in Table 1, the difference between the in-field linear errors within the same wafer is negligibly small. As shown by this result, it is possible to simultaneously perform the in-field linear error correction for each of the other exposure fields by only performing the alignment correction for one shot or several shots in the same wafer.

Example 2 Table 2 shows that four wafers (substrate numbers 1, 10, 20, and 30) belonging to the same lot were subjected to alignment by one-point measurement in the field using an i-line stepper, and then exposed. The error is measured for the substrate after exposure. Therefore, the in-field linear error component is a measured value when no alignment correction is performed.

[0047]

[Table 2]

As shown by the field magnifications and field rotations actually measured in Table 2, the difference (variation) in the field linear error between wafers belonging to the same lot is negligibly small. That is, the field magnification and the field rotation, which are the linear error components in the field, are so small that the fluctuation amount can be ignored in the actually measured value and theoretically within the same lot. Therefore, based on this result, in the case of a plurality of wafers belonging to the same lot, the in-field component alignment correction of the exposure field is performed every one or several wafers, and then this correction amount is set to the same lot. By applying the in-field component alignment of the exposure field of each of the other belonging wafers, the in-field linear error correction of each of the other wafers becomes possible.

Embodiment 3 Next, an embodiment relating to superposition accuracy control will be described in order below. First, in Example 3, four types of mechanical samples OL-1, OL-2, OL-3, and OL-4 were created. The OL-1 is formed by patterning the first-layer Al wiring film by i-line lithography and dry etching and then performing TE three times.
OS-CVD, resist coating, SOG coating,
This is a reproduction of the process between flattening the insulating film including etch back, and the measurement item is the overlay accuracy between the Al wiring and the contact hole. OL-2 is a reproduction of the process from the formation of the interlayer insulating film to the formation of the upper layer wiring, and the measurement items are contact holes and Al.
It is the overlay accuracy with the wiring. OL-3 is a reproduction of the process from the formation of a field oxide film by the polysilicon pad LOCOS method to the formation of the first W-polycide film through gate oxidation, and the measurement items are the field oxide film and the gate. It is the overlay accuracy with the electrodes. Furthermore, OL-4 reproduces the process from the patterning of the first layer W-polycide film by i-line lithography and dry etching to the formation of the LDD sidewall, and the measurement item is between the gate electrode and the node contact. Is the overlay accuracy of. The field size of each mechanical sample is 20 mm x 20 m
m, the design rule is 0.5 μm.

The sampling method is as shown in FIG. That is, three wafers W are sampled from one lot, four corners are selected from four exposure fields F arranged in a matrix of 5 × 5 on each wafer W, and four points are midpoints on one side. And a total of 9 exposure fields F in the center of the wafer are sampled, and 5 points are measured in one of the exposure fields F in the center of the wafer and 8 in the remaining peripheral area
In each exposure field F, one point measurement was performed. That is, the total number of measurement points is 13. The measurement point P in each exposure field F is shown by a black dot, and in the case of 1-point measurement, the field center was added, and in the case of 5-point measurement, four corners of the field were added.

By analyzing the measurement result of each mechanical sample, 3σ
Was calculated and summarized in Table 3. For comparison, as shown in FIG. 5 described above, the analysis result of 3σ by the conventional method (total number of measurement points = 45) of measuring five points in the field in all of the sampled exposure fields, and the book The difference between the invention and the conventional method is also shown.

[0052]

[Table 3]

From Table 3, it is clear that the results of the present invention and the results of the conventional method are in good agreement, and the validity of the present invention was proved. In the present invention, the number of 5-point measurement fields is set to be 1 for each wafer, whereby the time required for measurement is reduced by about 70% as compared with the conventional method.

In other words, the third embodiment corresponds to the case where the number of exposure fields for multipoint measurement is reduced for each substrate, and n samples (n is a natural number of 2 or more) sampled from one substrate.
In the exposure field of, the multi-point measurement in the field is performed for m (m is a natural number satisfying 1 ≦ m ≦ n−1), and the single point measurement is performed for the remaining (nm). . At this time, even if the number of multipoint measurement fields is one (m = 1) as in the present embodiment, it is sufficiently practical.

Embodiment 4 In the present invention, a modification as shown in FIG. 3 is possible.
That is, when, for example, three wafers W are sampled from one lot, five in-field measurements are performed on all nine exposed fields F sampled for one wafer, and nine exposed fields are used for the remaining two wafers. 1 point measurement in the field is performed for all of F. By reducing the number of multi-point measurement fields on a lot-by-lot basis as described above, the time required for measurement can be shortened by about 50% as compared with the conventional method.

That is, Example 4 corresponds to a case where the number of multi-point measurement exposure fields is reduced in lot units, and M (N is a natural number of 2 or more) substrates sampled from the lot are M ( M is a natural number satisfying 1 ≦ M ≦ N−1. In-field multipoint measurement is performed for a plurality of exposure fields sampled from the substrate, and for the remaining N−M substrates, 1 point in the field. It is to measure. In addition, when performing the intra-field error measurement for each board belonging to the same lot, if the intra-field error measurement result already obtained by any board belonging to this lot is used for other boards, The time required for measurement can be further shortened.

Embodiment 5 The present invention can also be modified as shown in FIG. That is, when, for example, three wafers W are sampled from one lot, only one of the nine exposure fields F sampled for one wafer is measured at five points in the field and the remaining eight wafers are measured at one field in the field. Point measurement is performed, and for the remaining two sheets, one field measurement is performed for all nine exposure fields F. By reducing the number of multi-point measurement fields in both lot units and substrate units in this way, the time required for measurement is about 80 compared with the conventional method.
% Can be shortened.

That is, in the fifth embodiment, with respect to M substrates, only a part of a plurality of sampled exposure fields is set as a multipoint measurement field, and the number of multipoint measurement fields is set in both lot units and substrate units. It corresponds to the case of reducing with. At this time, even if the number of substrates including the multipoint measurement field is one (M = 1) as in the present embodiment, it is sufficiently practical.

As described above, according to the present invention, it is possible to reduce the number of exposure fields for multipoint measurement at least on a substrate (wafer) basis or a lot basis.

In the present invention, the multi-point measurement field may be basically selected from any position on the substrate, but in order to suppress variations in each exposure field as much as possible, it may be selected with some symmetry. desirable. In particular, when the number of exposure fields for performing multipoint measurement is one, it is preferable to select this exposure field from the center of the substrate, whereby the influence of thermal deformation of the wafer can be minimized. .

Specific examples of the present invention have been described above, but the present invention is not limited to these examples. It goes without saying that details such as design rules, layout of exposure fields on the wafer, field size, sampling layout, number and layout of measurement points in the field can be appropriately changed and selected.

[0062]

As is apparent from the above description, according to the present invention, even if the number of multi-point measurement fields is reduced, practically sufficient accuracy of alignment, correction exposure, and post-exposure overlay are achieved. The accuracy can be controlled. Therefore, the number of samplings can be reduced without deteriorating the alignment accuracy (correction accuracy), the TAT can be improved, and the total time required for lithography can be shortened. As described above, the present invention greatly contributes to reliability improvement, cost reduction, and throughput improvement in semiconductor device manufacturing.

[Brief description of drawings]

FIG. 1 is a schematic diagram for explaining the definition of overlay error in an exposure field.

FIG. 2 is a schematic diagram showing an example in which the present invention is applied to reduce the number of multi-point measurement fields in substrate units.

FIG. 3 is a schematic diagram showing an example in which the present invention is applied to reduce the number of multipoint measurement fields in lot units.

FIG. 4 is a schematic diagram showing an example in which the present invention is applied and the number of multipoint measurement fields is reduced in both substrate units and lot units.

FIG. 5 is a schematic diagram showing a conventional method of performing 5-point in-field measurement on all of the sampled exposure fields for all of the wafers sampled from one lot.

[Explanation of symbols]

 W: wafer, F: exposure field, P: measurement point

Claims (13)

[Claims] 1. A method of measuring an alignment error that occurs in a plurality of exposure fields formed on a substrate in response to repeated lithographic shots on the substrate in a semiconductor device manufacturing process, wherein n is sampled from the substrate. M (n is a natural number of 2 or more) exposure fields, m (m is 1 ≦ m ≦ n−
The alignment error measuring method is characterized in that a multipoint measurement within the field is performed for a natural number satisfying 1, and a single point measurement within the field is performed for the remaining (nm) numbers. 2. The alignment error measuring method according to claim 1, wherein the m is 1. 3. A method of measuring an alignment error which occurs in a plurality of exposure fields formed on a substrate corresponding to repeated lithographic shots on the substrate in a semiconductor device manufacturing process, wherein a lot to which the substrate belongs is constituted. For M (M is a natural number satisfying 1 ≦ M ≦ N−1) substrates out of N substrates (N is a natural number of 2 or more), a plurality of exposure fields sampled from each substrate are used. In-field multi-point measurement is performed on at least one substrate, and for the remaining (NM) substrates, one-field measurement is performed on a plurality of exposure fields sampled from each substrate. How to measure alignment error. 4. The alignment error measuring method according to claim 3, wherein multi-point measurement in a field is performed for all of a plurality of exposure fields sampled for the M substrates. 5. An exposure method in which alignment is performed based on measurement of a predetermined portion of an exposure field formed on a substrate corresponding to repeated lithographic shots on the substrate in a semiconductor device manufacturing process. Of n (n is a natural number of 2 or more) exposure fields sampled from m (m is 1 ≦ m ≦ n−
1 field measurement for the remaining (n−m) fields and at least one exposure field field based on the one field multi-field measurement result. An exposure method, wherein internal component alignment is performed. 6. The exposure method according to claim 5, wherein the m is 1. 7. An exposure method in which alignment is performed based on measurement of a predetermined portion of an exposure field formed on a plurality of substrates in response to repeated lithographic shots on the substrate in a semiconductor device manufacturing process. Of all N substrates (N is a natural number of 2 or more) that compose the lot to which M number of substrates (M is a natural number satisfying 1 ≦ M ≦ N−1) including the first substrate of the lot In-field multipoint measurement is performed on at least one of a plurality of exposure fields sampled on each of the substrates, and a plurality of exposures sampled from each of the remaining (N−M) substrates. The field is measured at one point in the field, and based on the result of the multi-point measurement in the field of the M substrates, the in-field measurement is performed for the exposure field of the substrate. Exposure in which the alignment is performed and the intra-field multi-point measurement results made on the M substrates are applied to other (NM) substrates to perform the intra-field component alignment of the exposure field. Method. 8. The exposure method according to claim 7, wherein multi-field measurement is performed for all of the plurality of exposure fields sampled for the M substrates. 9. A method for measuring overlay error occurring in a plurality of exposure fields formed on a substrate corresponding to repeated lithography shots on the substrate, and managing overlay accuracy based on the measurement result. In the measurement of the overlay error, among n (n is a natural number of 2 or more) exposure fields sampled from the substrate, m (m is a natural number satisfying 1 ≦ m ≦ n−1) fields Multi-point measurement is performed and left (nm)
An overlay accuracy control method characterized by performing one point measurement in the field for individual pieces. 10. The overlay accuracy management method according to claim 9, wherein the m is 1. 11. A method of measuring overlay error occurring in a plurality of exposure fields formed on a substrate corresponding to repeated lithography shots on the substrate, and managing overlay accuracy based on the measurement result. , N pieces sampled from the lot to which the substrate belongs in the measurement of the overlay error (N is a natural number of 2 or more)
M substrates (M is a natural number that satisfies 1 ≦ M ≦ N−1)
In-field multipoint measurement is performed on at least one of the plurality of exposure fields sampled from each substrate, and the remaining (N−M) substrates are sampled from each substrate. An overlay accuracy management method characterized by performing one point measurement in each exposure field. 12. The overlay accuracy management method according to claim 11, wherein multipoint measurement is performed for all of the plurality of exposure fields sampled for the M substrates. 13. The overlay accuracy management method according to claim 11, wherein the M is 1.