JPH1055949A - Positioning method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a stable positioning accuracy with keeping a high throughput by monitoring the variation of error parameters in wafers or between wafers to dynamically change the number of measuring points. SOLUTION: Sno wafers are measured at a max. number of measuring points Mmax by the EGA method, beginning the top of a lot. Criterion items of following exposure process are selected to average error parameters obtained from exposed wafers and store them. S21 detects the positions of pattern regions at a min. no. of measuring points Mmin for the Nno+1-th wafer and which follow, S22 estimates the error parameters, if S23 judges them to be within approximate ranges, it obtains the shot position to execute the exposure process, using them. If outside the ranges, a position detected value of other pattern region is added to estimate the error parameters and this is repeated until the parameters are within the ranges to obtain the shot position to execute the exposure process.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an alignment method suitable for a projection exposure apparatus used for manufacturing a semiconductor device, a liquid crystal display device, or a thin film magnetic head, and more particularly, to a method for continuously processing a large number of substrates. The alignment method.

[0002]

2. Description of the Related Art In recent years, the degree of integration of semiconductor devices has increased dramatically, and the development of memories of 64 Mbits and 256 Mbits,
Production is underway. Further, in the liquid crystal display device, as the display area becomes larger and higher definition, the display pixels are becoming finer. As an exposure apparatus for manufacturing these semiconductor devices or liquid crystal display devices, a step-and-repeat projection exposure apparatus (hereinafter referred to as a stepper) is used.
Is used. The stepper performs exposure by aligning a projected image of a circuit pattern on a photomask or a reticle (hereinafter, collectively referred to as a reticle) for each local region on a semiconductor wafer or a glass substrate (hereinafter, collectively referred to as a wafer). The following four types are known as methods for performing this alignment with high accuracy.

First, as a first alignment method, there is a so-called die-by-die alignment method that optimizes the state of superposition of a projected image and a shot area for each exposure.
An observation optical system for positioning includes an on-axis method for simultaneously observing a positioning mark attached to a shot area where an exposure pattern on a wafer is formed and a positioning mark on a reticle, and a method of detecting a mark on a reticle. There is an off-axis method in which a mark on a wafer is detected by a mark detection system having a detection center at a position away from the projection point by a certain distance, and the wafer is fed by a certain distance from the detected position.

When die-by-die alignment is performed for each local region on a wafer, the alignment accuracy of each exposure shot is improved, but the processing time for one wafer is increased due to the large number of measurements. Or, if the mark formed in association with each shot area is deformed for some reason, there is a drawback that the exposure of the shot becomes an overlay failure.

As a second alignment method, alignment of the entire wafer with respect to the reticle is performed only once before exposure, and thereafter mechanical alignment is performed in a step-and-repeat manner in accordance with a design value of a shot arrangement on the wafer. ,
There is a so-called global alignment method. In the global alignment method, the position is measured by the mark coordinates (X, Y) of one point on the wafer, and the rotation of the wafer is measured by the X or Y coordinate value of another point, and the alignment error between the reticle and the wafer is measured. The shot arrangement on the wafer is determined based on the result. This method merely recognizes an error amount in only two directions, that is, an X direction and a Y direction, of a wafer stage as an error parameter (offset) by a simple averaging, and a magnification error (scaling) including expansion and contraction of the wafer itself. No consideration is given to errors due to rotation errors (rotation) of the wafer stage on the stage, orthogonality of running of the wafer stage, and the like. Therefore, although the time required for measurement is short in the global alignment method, the throughput (the number of processed wafers per unit time) can be improved, but it is difficult to obtain a high overlay.

A third alignment method is disclosed in
An alignment method disclosed in Japanese Patent No. 44429 will be described. First, the position of alignment marks in a plurality of (eg, 5 to 10) shot areas (sample shot areas) on a wafer designated in advance is measured, and scaling, rotation, offset amount, and offset amount are determined based on the actually measured alignment mark positions. The orthogonality of the wafer stage is determined using the linear least squares method. An equation for the regularity of the shot arrangement is determined using these four error components as error parameters, the shot position on the wafer is estimated based on the equation, and the wafer stage is positioned by the step-and-repeat method. This alignment method will be referred to as an enhanced global alignment (EGA) method.

In the EGA method, it is necessary to accurately obtain the above four error parameters. Therefore, alignment mark positions are measured for as many sample shot areas as possible on one wafer, and the actual measured values of the sample shot positions are obtained. It is necessary to obtain many. Usually, in a stepper, since a large number of wafers are continuously exposed in lot units, when the alignment accuracy of each wafer is increased by the EGA method, the number of sample shot areas in the wafer in the lot increases. Therefore, the number of times of alignment within one wafer increases,
As a result, there is a problem that the throughput is limited.

A fourth alignment method is disclosed in Japanese Patent Application Laid-Open No. 62-8 / 1987.
This is an alignment method disclosed in Japanese Patent No. 4516, which is an improvement of the above EGA method in lot-by-lot processing. A plurality of error parameters (scaling, orthogonality, etc.) for determining the regularity of the shot arrangement are determined for each of the m wafers (m <N) from the start of the processing for the N wafers in the lot. , Rotation, and shift) are estimated based on a predetermined number of actually measured values for each wafer. For each of the (m + 1) th and subsequent wafers, the number of measurement points is reduced, and among the plurality of error parameters, only error parameters (rotation and shift) whose values can change for each wafer are determined by a predetermined number (m-th ) And other error parameters (scaling and orthogonality)
With regard to (2), the error parameters already estimated for the first to m-th wafers are applied, and the formula of the regularity of the array is determined based on the estimated error parameters. In this way, the alignment of the first to Nth substrates is sequentially performed. As described above, according to Japanese Patent Application Laid-Open No. 62-84516, throughput can be improved while maintaining high alignment accuracy by appropriately setting the number of measurement points and processing wafers with stable lot conditions.

[0009]

However, the fourth alignment method is based on the premise that the in-lot conditions are stable, so that if the in-lot conditions become unstable for some reason, the error parameter can be accurately obtained. And the alignment accuracy may be reduced. In order to keep the alignment accuracy constant, redundancy is given to the number of measurement points per wafer and the number of measurement wafers for estimating error parameters in anticipation of variations in lot conditions, If a 100% inspection is performed, the throughput cannot be improved.

As shown in FIG. 8, the vertical axis represents the variation of the error parameter and the processing time, and the horizontal axis represents the number of measurement points on the wafer. As the number of measurement points increases, the processing time becomes linear as indicated by the broken line. Increase. On the other hand, the variation of the error parameter (solid line) increases at an accelerated rate as the number of measurement points decreases. As described above, since the improvement of the alignment accuracy and the improvement of the throughput are in a contradictory relationship, the position at which the optimum alignment can be performed according to various conditions such as the environment at the time of projection exposure and the required performance for the semiconductor element to be formed. A matching method is desired.

The present invention provides a positioning method which improves throughput while maintaining high positioning accuracy for each pattern (shot area) in each wafer when a plurality of wafers are continuously aligned. The purpose is to do.

[0012]

SUMMARY OF THE INVENTION The object of the present invention is to provide an image processing method for each of N substrates on which a plurality of pattern areas are formed, based on a design value relating to the position of the pattern area and a plurality of error parameters. In a positioning method for determining and positioning a shot position to be actually exposed, the positions of Mmax pattern regions are set for each of the first to m-th (m <N) substrates. A first step of estimating all of a plurality of error parameters based on a first actually measured value obtained by measuring the position and determining a shot position to be actually exposed; , One or more predetermined error parameters are selected, and the respective values of the selected error parameters obtained from each of the m substrates are averaged, respectively, to register the registered first The second step to be stored in each criterion item in the “criterion”, and for each of the (m + 1) th and subsequent substrates, the number of measurement points is obtained by measuring the positions of Mmin (<Mmax) pattern areas. If the value of the error parameter selected from all of the plurality of error parameters estimated based on the second actually measured value is within a predetermined range with respect to the first “judgment criterion”, the first “judgment” A third step of determining a shot position to be actually exposed based on an error parameter stored in each determination reference item in the “reference” and other error parameters obtained by the actual measurement value;
If the value of the selected error parameter among the plurality of error parameters estimated based on the second actually measured value is not within the predetermined range, the number of measurement points is increased until the value of the selected error parameter falls within the predetermined range. And a fourth step of additionally performing re-measurement.

It is another object of the present invention to provide a positioning method in which the value of the selected error parameter estimated during the fourth step has a tendency different from the first "judgment criterion". When the change in the value of the selected error parameter decreases and shows a predetermined convergence tendency even when the score is increased, the value of the error parameter selected in each of the first criterion items is updated. A fifth step of creating a new second "criterion" and additionally registering the error parameter stored in each criterion item in the second "criterion" and measuring in the fifth step A sixth step of determining a shot position to be actually exposed based on the obtained second actually measured value and other error parameters obtained from the actually measured value based on the added number of measurement points. To do It is achieved by the combined method.

It is still another object of the present invention to determine a shot position to be actually exposed based on the second "judgment criterion" and other error parameters obtained by new actually measured values. And 7 steps.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A positioning method according to an embodiment of the present invention will be described with reference to FIGS. 1 to 4 are diagrams for explaining the processing flow of the positioning method according to the present embodiment. FIG. 5 is a diagram showing a schematic configuration of a stepper for performing the alignment method according to the present embodiment. FIG. 6 shows a pattern (shot area) arrangement on a wafer processed by the stepper of FIG. FIG.

First, a schematic configuration of a stepper according to the present embodiment will be described with reference to FIG. Reticle R on which predetermined pattern area (including alignment mark) PA is formed is accurately positioned with respect to optical axis AX of projection lens 1. The projection image of the pattern PA is transferred to the wafer W mounted on the stage 2 that moves two-dimensionally in the X and Y directions. 11 denotes a principal ray passing through the outermost edge of the pattern area PA. In the present embodiment, it is assumed that the object (reticle R) side of the projection lens 1 is a non-telecentric system. The stage 2 is driven by a motor 3, and its two-dimensional position (coordinate value) is measured by a laser light interferometer 4.

In order to detect alignment marks (particularly diffraction grating marks) formed in advance on the wafer W,
A laser light source 5, a half mirror 6, and mirrors 7 and 8 for outputting a laser light, such as He-Ne, which hardly exposes the photoresist on the wafer, is provided. The laser light from the laser light source 5 passes through the projection lens 1. To form an image on the wafer W as a spot light SP. The spot light SP is arranged on the wafer W outside the projected image of the pattern area PA so as to be located at a fixed distance from the optical axis AX. When the spot light SP irradiates the mark, diffracted light, scattered light, and specularly reflected light are generated.
7 and a spatial filter 9 via the half mirror 6. The spatial filter 9 is arranged conjugate with the entrance pupil of the projection lens 1, and is formed so as to cut the zero-order light (specular reflection light) and guide the diffracted light (or scattered light) to the photoelectric detector 10. The position of the mark on the wafer W is detected by the spot light SP.
Is fixed in the projection field of view of the projection lens 1, and is executed by the main control circuit 11 which receives the coordinate value from the interferometer 4 for detecting the moving position of the stage 2 and the photoelectric signal from the photoelectric detector 10. Is done. The main control circuit 11 further controls the driving of the motor 3, and performs a step-and-repeat exposure operation by the EGA method and an operation of determining the regularity of the arrangement of the shot areas on the wafer W (main operation operation of the EGA method). )I do. The main control circuit 11 also stores the design values of the shot arrangement in advance.

The alignment system including the laser light source 5, the half mirror 6, the mirrors 7, 8, the spatial filter 9, and the photoelectric detector 10 will be hereinafter referred to as a laser step alignment system (LSA system). The detection center of the LSA system is the center of the spot light SP. The LSA system shown in FIG. 5 is for detecting only the position of the wafer W in the Y direction, for example, and the LSA system for detecting the position in the X direction is actually arranged similarly. FIG. 5 shows only the first mirror 8 ′ of the LSA system for X direction detection corresponding to the first mirror 8 of the LSA system for Y direction detection.

Now, the wafer W placed on the stage 2
Above, as shown in FIG. 6, a plurality of rectangular pattern areas C
P is formed in a matrix along the arrangement coordinate system aβ. Each of the pattern areas CP is determined so as to overlap the projected image of the pattern area PA of the reticle R, and each pattern area CP has an X-direction alignment mark SX and a Y-direction alignment mark SY. It is formed concomitantly. Set the origin of the array coordinate system αβ to the wafer W
If the pattern area CP0 is determined to coincide with the center point of the pattern area CP0 located near the upper center, the pattern area CP0
0 is located on the α axis, and the mark S
X0 is formed so as to be located on the β axis. The marks SY and SX attached to the other pattern areas CP are formed according to the same rule. Design coordinate values (or X direction and Y direction) of each pattern area CP in the array coordinate system αβ
The direction stepping pitch) is stored in the main control circuit 11 in FIG. 5 in advance.

FIG. 7 shows a pattern area CP by the LSA system.
FIG. 4 is a plan view showing an arrangement relationship between a spot light SP and a wafer W when detecting the position of the wafer W. X of rectangular coordinate system X, Y
The axis and the Y axis represent the moving direction of the stage 2 (or the direction of coordinate measurement by the interferometer 4). Here, the origin of the coordinate system X, Y coincides with the center of the projection field IF of the projection lens 1 (optical axis AX). It is defined as follows. LS for alignment in Y direction
Spot light SP by A system (hereinafter referred to as Y-LSA system)
Y is formed in a strip extending elongated on the X axis, and the spot light SPX by the LSA system for alignment in the X direction (hereinafter referred to as X-LSA system) is formed in a strip elongated in the Y axis.

A rectangular area indicated by a broken line in FIG. 7 is a projection image PA 'of the pattern area PA of the reticle R, and the projection image P
The center of A '(the center of the shot) is assumed to coincide with the origin of the coordinate system XY. The marks SY and SX attached to the pattern area CP on the wafer W are diffraction grating patterns. Then, the wafer W (stage 2) is moved so that the mark SY crosses the spot light SPY in the Y direction, the position YYi of the diffracted light generated from the mark SY in the Y direction is read from the interferometer 4, and the mark SX is X The wafer W (stage 2) is moved so as to traverse the spot light SPX in the direction, and the position XXi in the X direction of the diffracted light generated from the mark SX is read from the interferometer 4 to obtain the measured value of the position of the pattern area CP. (XXi, YYi) is obtained. Therefore, the read value of the interferometer 4 is the measured value (XXi, YY).
By positioning the stage 2 so as to coincide with i), the center CC of the pattern area CP and the center (origin) of the projection image PA ′ exactly match, and the pattern area CP and the projection image PA ′ Will be superimposed precisely. In the present embodiment, since the LSA system does not perform alignment and exposure for each pattern area CP, but rather uses the EGA method, the actual measurement values of the positions of the marks SX and SY are obtained. Operation, so-called sample alignment, may be performed.

Before describing the positioning method according to the present embodiment, the EGA method described in Japanese Patent Application Laid-Open No. 61-44429 will be briefly described. Regarding the regularity of the arrangement of shots (pattern regions CP) on one wafer, the following six error parameter variables are introduced assuming linear distortion on a plane. Mx: magnification error in wafer X direction (scaling X) My: magnification error in wafer Y direction (scaling Y) ：: rotation error (rotation) of array coordinate system αβ ω: orthogonality of coordinate system XY Sx: wafer The amount of translation in the X direction (shift X) Sy: The amount of translation in the Y direction of the wafer (shift Y) Assuming the above variables, the design coordinate values (Xon,
The coordinate value (Xrn, Yrn) of the shot position to be actually positioned of the shot located at (Yon) is expressed as follows. Here, n indicates a shot number.

[0023]

(Equation 1)

The measured value in the actual alignment is (Xr
n ′, Yrn ′), the error component Exn in the X direction is Exn = Xrn′−Xrn, and the error component Eyn in the Y direction is Eyn = Yrn ′.
−Yrn, and these sums of squares are as follows.

[0025]

(Equation 2)

The above-mentioned six error parameters for minimizing the error E are obtained by the least-squares method, whereby the coordinate values (Xrn, Yrn) of the actual shot position with respect to the design coordinate values (Xon, Yon) of each shot. Can be uniquely determined.

The processing flow of the positioning method according to the present embodiment will be described with reference to FIGS. First,
One of N (for example, N = 20) wafers W to be subjected to exposure processing as a lot unit is pre-aligned using an orientation flat of the wafer W and loaded into a stepper, and is loaded into a stepper wafer stage 2. It is placed on top (Step S10). Next, after the rotation of the wafer W is corrected, the mark S attached to the pattern area CP on the predetermined wafer W by the X-LSA system and the Y-LSA system.
The positions of Y and SX are detected. The stage 2 is sequentially moved to detect the marks of the predetermined pattern area CP corresponding to the preset maximum number of measurement points (Mmax) (step S1).
1). The maximum number of measurement points Mmax is desirably about 10 shots, but the number of shots can be reduced in consideration of throughput.

Next, Mmax actual measured coordinate values (Xr1 ', Yr1') to (XrMm) obtained by the mark detection in step S11.
ax ', YrMmax') and design coordinate values (Xo1,
The above six error parameters are estimated using the least squares method from (Yo1) to (XoMmax, YoMmax) (step S).
12). Then, a shot position (Xrn, Yrn) to be actually exposed is calculated from the obtained error parameter and the design coordinate value using the above equation, and an exposure process is performed on the wafer W (step S13).

The wafer W after the exposure processing is placed on the stage 2
It is determined whether or not the process has been performed for a predetermined number of sampled wafers (Sno) (step S15). This Sno is a preset value of the number of wafers from the first wafer W of the lot.
Until the processing of one wafer is completed, the measurement of the maximum number of measurement points Mmax is performed on each wafer W by the conventional EGA method. Sno is set in consideration of the stability of the wafer (lot condition).

When the number of processed wafers reaches a predetermined number of Sno, the process proceeds to step S16, and a "judgment criterion" used to reduce the number of measurement points in the exposure processing of the Sno + 1th and subsequent wafers is created. The criterion is composed of one or more criterion items. One criterion item is 6
One of the two error parameters is selected. As the value of the error parameter selected as the judgment criterion item, a value obtained by averaging the error parameters obtained from each of the Sno wafers W that have already been exposed is stored.

The number of judgment criteria items of "judgment criteria" is arbitrary within a range of 1 to 6 depending on how many of the six error parameters are selected. Parameters should be selected as criteria. For example, magnification errors (scaling X, scaling Y) and orthogonality error parameters between wafers in a lot for which continuous processing is performed have extremely small variations, and thus are obtained by measuring each of the first to Sno wafers W. The average value of each of the three error parameters may be selected as a criterion item.

The small variations in the scaling X, Y and orthogonality are that the wafers in a lot are almost equally affected in various processes, and that the distortion of the projection optical system of the stepper and the orthogonality of the stage are different. The degree depends on the fact that it does not greatly change during lot processing. On the other hand, the mechanical operation given to each wafer when it is carried into the stepper and placed on the stage is slightly different for each wafer, so that it is easy to cause the wafer W to be attracted, and the shift X, Y, rotation The error does not reduce the variation.

For the reasons described above, in the present embodiment, three items of scaling X, scaling Y and orthogonality are used as judgment criteria items.

The "judgment criteria" may be newly created separately in the subsequent processing, and will be denoted by numbers to distinguish them. Here, since it is the first criterion created, it is referred to as criterion (1) hereinafter.

After the determination criterion (1) is set, the process proceeds to the exposure processing for the remaining Sno + 1-th and subsequent wafers W. In step S20, new wafers W after the (Sno + 1) th wafer are carried into the stepper, pre-aligned, and placed on the wafer stage of the stepper. Next, after the rotation of the wafer W is corrected, the X-LSA system and the Y-LSA
The system detects the positions of the marks SY and SX attached to the pattern area CP on the predetermined wafer W. The stage is sequentially moved, and the preset minimum number of measurement points (Mmi
Mark detection of the predetermined pattern area CP for n) is performed (step S21). Since the minimum number of measurement points Mmin is sufficient for at least three shots (marks SX and SY can be detected in each shot) in order to obtain six error parameters, Mmin = 3 in this embodiment.

Next, the measured coordinate values (Xr1, Yr1) to (Xr3, Yr1) obtained by the mark detection in step S21.
From 3) and the design coordinate values stored in advance, an error parameter is estimated using the least squares method (step S22).

The values of the error parameters (scaling X, Y and orthogonality) stored in each of the three determination criteria of the criterion (1) and the values of the scaling X, Y and orthogonality obtained in step S22 It is determined whether the value of the error parameter is within the range of the approximation determination value (Jlevel) (step S23). The approximate judgment value is
A threshold value which is provided corresponding to each judgment criterion item and determines whether or not the values of the measured magnification errors X and Y and the error parameter of the orthogonality are within a predetermined allowable range of each judgment criterion item. (Mlx (unit ppm): scaling X, Mly (unit ppm): scaling Y and O
1 (unit μrad: orthogonality) is set. These approximate determination values are set in consideration of the stability of the measurement sensor and the wafer (lot condition).

If the value corresponding to the criterion item of the error parameter obtained in step S22 is within the approximate criterion value, the process proceeds to step S42, where the error parameter of the criterion (1) (scaling X, Y And orthogonality) and other error parameters (shifts X, Y, and rotation) obtained from the actual measurement of the wafer W, and a shot position to be actually exposed is calculated from the above equation and the design coordinate value using the above equation. The exposure process is performed on the wafer W by moving the stage 2 by a predetermined amount based on the calculation result (step S42). The exposed wafer W is unloaded from the stage 2 (step S43), and it is determined whether the exposure processing for all the wafers W in the lot has been completed. If there is an unprocessed wafer W, the process returns to step S20 to return to the unprocessed wafer W The wafer W is exposed (step S44).

On the other hand, if it is determined in step S23 that the value corresponding to the obtained criterion item of the error parameter is out of the approximate determination value, the X-LSA system and the Y-LS
The positions of the marks SY and SX attached to yet another predetermined pattern area CP on the wafer W are detected using the A system (step S30). Then, the actually measured coordinate values (Xr4, Yr4) newly obtained in step 30 and the actually measured coordinate values (Xr1, Yr4) already obtained by the mark detection in step S21.
Error parameters are estimated again by the least squares method using the four data of 1) to (Xr3, Yr3) (step S31).

Then, again, the error parameters (scaling X, Y and orthogonality) of the criterion item of the criterion (1)
Then, it is determined whether or not those values of the error parameter obtained in step S31 are within the range of the approximate determination value (Jlevel) (step S32). If the value corresponding to the criterion item of the error parameter obtained in step S31 is within the approximate criterion value, the process proceeds to step S42 and the error stored in each criterion item of the criterion (1) is determined. Using parameters (scaling X, Y, and orthogonality) and other error parameters (shift X, Y, and rotation) obtained by actual measurement of the wafer as an error parameter, and using the above equation from the design parameter and the error parameter A shot position to be actually exposed is calculated, and an exposure process is performed on the wafer W. The exposed wafer W is unloaded from the stage (step S43), and it is determined whether the exposure processing of all the wafers in the lot has been completed. If there is an unprocessed wafer, the flow returns to step S20 to perform the wafer exposure processing. Perform (Step S44).

On the other hand, even if the number of measurement points is increased by one, the error parameters (scaling X, Y
If the respective values of (及 び and the orthogonality) do not fall within the respective approximate determination values (step S32), the process proceeds to step 33, and the variation of the error parameter obtained in step S31 is within the range of the convergence determination value. It is determined whether or not. The convergence criterion value is set for each criterion item, and a threshold value (M
wx (unit: ppm): scaling X, Mwy (unit: p)
pm): Scaling Y, and Ow (unit μrad):
Orthogonality) are set respectively. These convergence determination values are set in consideration of the stability of the measurement sensor.

If each of the error parameters (scaling X, Y and orthogonality) obtained in step S31 is not within the range of the convergence judgment value, the number of measurement points is set to Mm in step S34.
ax is determined, and if not, the process returns to step S30 to further newly select marks SX and SY of another pattern area CP and measure their positions, thereby increasing the number of measurement points by one. An error parameter is estimated based on the coordinate data (step S31). Step S34
If it is determined that the total number of measurement points exceeds Mmax, it is determined that there is some abnormality in the measurement sensor or the wafer, and the flow shifts to step S43 to end the exposure operation and carry the wafer out of the apparatus.

When the change amount of the error parameter falls within the range of the convergence judgment value, in step S40, the error parameter obtained from the wafer is changed to the judgment criterion set value with respect to the judgment criterion (1). It is determined whether or not (Jdelta) has been changed. Judgment standard setting value is
A threshold value (Mdx (unit: pp) for defining the error parameter obtained from the measurement result
m): scaling X, Mdy (unit ppm): scaling Y, and Od (unit μrad): orthogonality). These criterion setting values are set for each criterion item in consideration of the stability of the measurement sensor.

The error parameter obtained from the wafer corresponds to the criterion set value (Jdel) with respect to the criterion (1).
If the value exceeds ta), a new “judgment criterion” is created and used as the criterion (n). Here, (n) indicates that it is set as the n-th criterion. In the three criterion items of the criterion (n), the scaling X, Y and orthogonality of the six error parameters estimated in step S31 are selected, and the values of the respective error parameters are stored (step S31). S41).

The error parameters (scaling X, Y and orthogonality) stored in each criterion item of this criterion (n) and other error parameters (shift X, Y and rotation) obtained by actual measurement of the wafer And using
A shot position to be actually exposed is calculated from the design coordinate values by the above equation, and an exposure process is performed on the wafer W (step S42). The exposed wafer W is unloaded from the stage (step S43), and it is determined whether the exposure processing of all the wafers in the lot has been completed. If there is an unprocessed wafer, the flow returns to step S20 to perform the wafer exposure processing. Continue (step S44).

On the other hand, if the error parameter obtained from the wafer in step S40 does not exceed the criterion set value (Jdelta) with respect to the criterion (1), no new "criterion" is created. In this case, the error parameters (scaling X, Y and orthogonality) stored in each of the existing “criterion” criteria and other error parameters (shift X, Y and rotation) obtained by actual measurement of the wafer ), The shot position to be actually exposed is calculated from the design coordinate value by the above formula, and the exposure process is performed on the wafer W (step S42).

By doing so, the wafers W in the lot
All are exposed with proper alignment accuracy and wafer processing time. The above processing is summarized as follows.
For each wafer W from the first wafer to the Sno wafer in the lot processing, mark detection is performed on Mmax pattern areas CP at predetermined positions to estimate six types of error parameters. A shot position to be exposed is determined.

From among the estimated six types of error parameters, one or more predetermined error parameters are selected, and S
The value of the error parameter obtained from each of the no wafers is averaged, and stored in each of the criteria items in the registered “criteria”. This "criterion" is Sno +
It is used for error parameter estimation in the processing of the first and subsequent wafers, and has the effect of reducing the number of measurement points in the pattern area CP of the Sno + 1 and subsequent wafers.

The error parameter value estimated by performing mark detection on Mmin (<Mmax) pattern areas CP with a reduced number of measurement points for the Sno + 1th and subsequent wafers is determined by a predetermined value with respect to the “judgment criterion”. If it is within the range,
A shot position to be actually exposed is determined using the value of the error parameter stored in each of the criteria items in the “criterion” and other error parameters obtained by actual measurement of the wafer.

If the value of the error parameter estimated by reducing the number of measurement points is not within a predetermined range with respect to the “criterion”, the value of the estimated error parameter is within a predetermined range with respect to the “criterion”. Measurement is performed by adding the number of measurement points until entering. At this time, if the value of the estimated error parameter shows a tendency different from the “judgment criterion”, and the value of the error parameter shows a small change even when the number of measurement points to be added is increased, and shows a predetermined convergence tendency, The corresponding value of the error parameter is stored in each criterion item, a new “criterion” is created, and additionally registered.

A shot position to be actually exposed is determined using the value of the error parameter stored in each judgment criterion item in the new "judgment criterion" and other error parameters obtained by actual measurement of the wafer. I do.

In the subsequent measurement for the wafer, any one or more of the registered “criteria” or the “criterion” newly registered in the measurement and other errors obtained by the actual measurement of the wafer A shot position to be actually exposed is determined using the parameters.

As described above, according to the positioning method according to the present embodiment, the conventional EGA method is developed, and attention is paid to the fluctuation state of the error parameter within the wafer and between the wafers.
Since the number of measurement points of the actual measurement value for estimating the error parameter is dynamically changed based on the fluctuation of the value of the predetermined error parameter, a high throughput with a desired positioning accuracy can be obtained in a stable lot processing. Therefore, in an unstable lot process, it is possible to realize a stepper capable of obtaining stable alignment accuracy while suppressing a decrease in throughput.

Next, an outline of a control environment for implementing the positioning method according to the present embodiment will be described. The alignment method in the present embodiment exists as a part of the software system used in the main control circuit 11 for controlling the stepper shown in FIG. From the side of the operator who operates the stepper, various items related to lot processing are interactively provided by an engineering work station as an information input processing device attached to the stepper or an input display device (keyboard, mouse, display) of a mini computer. Processing condition parameters can be freely set and edited. The processing condition parameters include EG
Including mark coordinates (points) for executing method A,
Information about the entire exposure processing such as an exposure layout and a reticle layout is included. These pieces of information can be defined (managed) for each lot and can be set according to the lot-specific required performance. Further, these pieces of information are stored in a storage device in the form of a file, and can be referred to as needed.

As described above, since the various parameters used for the alignment method can be set by the operator himself, it is possible to flexibly cope with the different required performance for each lot processing, and to optimize the throughput and the position for each lot. The alignment accuracy can be obtained.

The present invention is not limited to the above embodiment, and various modifications are possible. For example, in the above-described embodiment, the number of judgment criteria items of “judgment criteria” is set to 3, and error parameters of scaling X, scaling Y, and orthogonality, which are small in variation among wafers in a lot, are respectively set as judgment criteria items. However, shifts X, Y and rotation of other error parameters can also be selected as judgment criteria items if the error of each wafer in a lot is reduced due to the improvement of the performance of the stepper. In this case, the throughput is further increased. Can be improved.

[0057]

As described above, according to the present invention, fluctuations in error parameters within a wafer and between wafers are monitored and the number of measurement points is dynamically changed. In this method, the number of measurement points is minimized, and stable alignment accuracy can be obtained while maintaining high throughput. In the case of processing a lot with unstable lot conditions, the number of measurement points is increased until the error parameter in the wafer becomes stable.
Although the throughput is slightly reduced, the exposure processing can be performed while maintaining stable alignment accuracy. Regarding a slight decrease in the throughput, the overall processing capacity is improved by taking into account the fact that the time required for performing the sampling inspection or the whole inspection as in the related art and reducing the lot reprocessing time is reduced.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a positioning method according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating a positioning method according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating a positioning method according to the embodiment of the present invention.

FIG. 4 is a diagram illustrating a positioning method according to the embodiment of the present invention.

FIG. 5 is a diagram illustrating a configuration of a stepper used in the alignment method according to the embodiment of the present invention.

FIG. 6 is a plan view showing a pattern arrangement on a wafer processed by the stepper of FIG. 5;

FIG. 7 is a plan view showing an arrangement relationship between a spot light SP and a wafer W when a position of a pattern area CP is detected by an LSA system.

FIG. 8 is a diagram showing the relationship between the number of measurement points in a wafer, alignment accuracy, and wafer processing time.

[Explanation of symbols]

 Reference Signs List 1 projection lens 2 stage 3 motor 4 laser light wave interferometer 5 laser light source 6 half mirror 7, 8 mirror 9 spatial filter 10 photoelectric detector 11 main control circuit

Claims (6)

[Claims] 1. A method according to claim 1, wherein each of the N substrates on which a plurality of pattern areas are formed is actually exposed to each of the pattern areas based on a design value relating to the position of the pattern area and a plurality of error parameters. In a positioning method for determining and positioning a shot position to be performed, for each of the first to m-th (m <N) substrates, Mmax positions of the pattern areas are measured. A first step of estimating all of the plurality of error parameters based on the obtained first actually measured values and determining a shot position to be actually exposed; Or 2
The above-mentioned predetermined error parameters are selected, the respective values of the selected error parameters obtained from each of the m substrates are averaged, and the values are registered in the registered first “judgment criterion”. A second step of storing in each determination criterion item, a second step obtained by measuring the positions of the pattern areas having Mmin (<Mmax) measurement points for each of the (m + 1) th and subsequent substrates. If the value of the selected error parameter among all of the plurality of error parameters estimated based on the actual measurement value is within a predetermined range with respect to the first “judgment criterion”, the first “judgment” A third step of determining a shot position to be actually exposed based on an error parameter stored in each determination reference item in the “reference” and other error parameters obtained from the actual measurement value; Measurement 2 If the value of the selected error parameter among the plurality of error parameters estimated based on the value is not within the predetermined range, the number of measurement points is added until the value of the selected error parameter falls within the predetermined range. And a fourth step of performing re-measurement. 2. The alignment method according to claim 1, wherein the value of the selected error parameter estimated during the fourth step has a tendency different from the first “criterion”, and If the change in the value of the selected error parameter decreases and shows a predetermined convergence tendency even when the number of measurement points to be added is increased, the error parameter selected for each of the first "judgment criteria" A fifth step of creating and additionally registering a new second “determination criterion” in which the value of is updated, and the error parameter stored in each criterion item in the second “criterion” and the A sixth step of determining a shot position to be actually exposed based on the second actually measured value measured in the fifth step and other error parameters obtained from the actually measured value based on the added number of measurement points. Step Alignment method characterized by comprising and. 3. A positioning method according to claim 2, wherein a shot position to be actually exposed is determined based on said second "judgment criterion" and other error parameters obtained by new measured values. And a seventh step of performing the alignment. 4. The positioning method according to claim 1, wherein the value of Mmax is 4 to 10. 5. The positioning method according to claim 1, wherein the value of Mmin is 3. 6. The alignment method according to claim 1, wherein the plurality of error parameters include a magnification error M of the substrate, a shift amount S of the substrate, and an array coordinate system of the pattern area. , The orthogonality ω of the array of the pattern areas
Wherein the magnification error M and the orthogonality ω are selected as the judgment reference items.