JP3223547B2 - Position control apparatus, exposure apparatus having the position control apparatus, device manufactured by the exposure apparatus, position control method, exposure method using the position control method, and device manufacturing method including the exposure method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a positioning apparatus for positioning one of a plurality of regions arranged on a substrate at a specific treatment position based on a design arrangement regularity. After calculating the target stop coordinate position of the movable stage based on the actual measurement value of the coordinate position of the area and the array coordinate value in design, the movable station is moved toward the target stop coordinate position.
A positioning device that moves and stops the laser, and several shot regions are selected from a large number of shot regions arranged on the substrate based on the arrangement regularity in the design, and the coordinate position is actually measured. Each target stop coordinate position of the movable stage corresponding to each of all shot areas (alignment position with the mask pattern) based on the actually measured value and the designed array coordinate value
And a moving stage is moved and stopped in a step-and-repeat manner toward each target stop coordinate position to thereby align the mask pattern with the shot area.

[0002]

2. Description of the Related Art A movable stage on which a sample is placed is moved to a specific processing position, and one of a plurality of regions regularly arranged on the surface of the sample is sent to the processing position, and this region is processed. After a two-dimensional alignment with respect to the position, a series of procedures of performing processing operations such as projection exposure and functional inspection on this area are repeated one after another, and as a result, processing is performed on all the areas on the sample. Various devices for completing the operation have been put to practical use. 1 on the sample in such a device
As an example of a two-dimensional alignment method between one area and a processing position, Japanese Patent Application Laid-Open No. 61-44429 discloses a method in which a large number of areas arranged on a substrate are arranged based on an arrangement regularity in design. Optionally select some regions to sample,
The coordinate positions of these sampling areas are actually measured, and the coordinate positions of all the areas on the substrate are calculated based on these actually measured values and the design coordinate values. By simply moving the movable stage toward the set coordinate position and stopping the movable stage, mechanical alignment can be performed without detecting any position of the area (or an alignment mark formed in association with the area). How to complete is introduced.

According to this method, since the respective error amounts of the detection results in several selected sampling regions are averaged, accurate positioning can be performed by utilizing the regularity of the arrangement of the regions on the substrate. It is. In addition, since the necessary positioning is completed only by the mechanical step-and-repeat progressive feeding, the position is detected independently in each area before the processing operation, and the positioning is repeated in all the areas. In comparison, the total processing time is reduced.

An alignment apparatus that aligns each area on a substrate to a processing position (for example, an exposure position) by using a method introduced in Japanese Patent Application Laid-Open No. 61-44429 discloses a method in which a plurality of areas are regularly aligned with each other. A movable stage capable of holding a three-dimensionally arranged substrate and moving along the arrangement direction of each region, coordinate measuring means for measuring a coordinate position of the movable stage, and a movable stage
Actual measuring means for measuring the coordinate positions of several areas on the substrate in advance in cooperation with the coordinate measuring means, and measuring the coordinate positions of these areas and the design of all the areas on the substrate. Calculating means for calculating the actual coordinate position (target position) of the entire area by performing a statistical calculation based on the array coordinate position of the movable stage, and the coordinate position of the movable stage measured by the coordinate measuring means every moment Stage control means for positioning the movable stage so that the calculated actual coordinate position (target position) is equal to the actual coordinate position (target position). After the calculation of the coordinate position (target position), the alignment between each area and the processing position is performed only by mechanical forward movement by the stage control means.

As an application example of such a positioning device,
A projection exposure apparatus (stepper) for sequentially sending and positioning a large number of exposure areas arranged on an exposure sample (photosensitive substrate) to a mask pattern image in a field of view of a projection optical system in a step-and-repeat manner can be used. . The projection exposure apparatus superimposes a mask pattern image and one of the exposure areas two-dimensionally within a projection image plane of a projection optical system, and then irradiates exposure light such as ultraviolet rays from the back side of the mask pattern. An apparatus for projecting a mask pattern onto an exposure area, wherein the mask pattern is transferred onto the exposure area on which a photosensitive material has been applied. Projection exposure apparatuses have been put to practical use in various fields of precision processing, such as the manufacture of semiconductor devices, and development research according to each application has been actively conducted. In particular, with the recent increase in the degree of integration of semiconductor devices and the reduction in design line width, projection exposure apparatuses for semiconductor manufacturing require a total alignment accuracy of which point within one exposure area. In this case, while the level reaches 0.1 μm or less, a decrease in throughput (the number of processed wafers per unit time) is not allowed, and an alignment technique (alignment of the exposure area and the mask pattern) that can achieve both accuracy and efficiency. ) Is strongly required.

[0006] A projection exposure apparatus for manufacturing semiconductors, which performs alignment of exposure areas using the method disclosed in Japanese Patent Application Laid-Open No. 61-44429, for example, requires a large number of exposure areas (also referred to as shot areas) to be regulated. An XY stage movable on the two-dimensionally arranged wafers along the arrangement direction of the exposure area and the coordinate position of the XY stage are measured in real time, and the measured values are measured. And a laser interferometer for outputting the coordinate position of the center of the field of view of the projection optical system on the XY stage, that is, the projection center point of the mask pattern, based on the XY stage. Measuring means for measuring the coordinate position of the exposure area in advance by detecting the alignment mark attached to the exposure area and detecting the coordinate position of the mark by a laser interferometer; and measuring the coordinate position of the exposure area in advance. Value and wafer Calculating means for calculating the actual coordinate positions of a plurality of exposure areas to be exposed by performing a statistical calculation based on the designed arrangement coordinate positions of the exposure areas in the above, and measuring every moment by a laser interferometer XY stay
X so that the coordinate position of the
And stage control means for positioning the Y stage, after the actual measurement means actually measures the coordinate positions of some exposure areas and the arithmetic means calculates the coordinate positions of all the areas. The XY stage is sequentially fed by the stage control means in a step-and-repeat manner, and the projection exposure at each stop position is repeated.

In the method disclosed in Japanese Patent Application Laid-Open No. 61-44429, the regularity of the actual arrangement of the regions on the substrate is determined by the relationship according to the design coordinate values or at most a simple proportional relationship (linear relationship). ), The coordinate positions of all the regions on the actual substrate are calculated, so that the regularity determined by each array coordinate value of the regions on the actual substrate is determined by design. If there is an error in the arrangement regularity, the positioning of each area by the stage control means becomes inaccurate. That is, when the regularity specified by the arrangement of each region on the substrate is equal to the arrangement regularity determined by design, all the regions on the substrate have linear position errors with respect to the design values. If there is no simple proportional relationship, the area is enlarged where the deviation between the actual array position of each area and the calculated coordinate position which is the stop target of the movable stage is not in a proportional relationship. Here, when the deviation amount between the actual arrangement position and the designed arrangement coordinate position varies randomly in each region, there is little room for further improving the alignment accuracy, but there is little room for improving the alignment accuracy. If a reproducible nonlinearity can be found in the distribution of the position deviation amount of the region, the final positioning accuracy can be further improved. For example, the designed array coordinate position is corrected using the nonlinearity of the deviation, and a corrected designed array coordinate position closer to the actual array state of the region on the substrate is obtained. By performing a statistical calculation based on the actual measured values of the coordinate positions of the sampling area and the corrected array coordinate positions on the design, the coordinate position of each area calculated even when there is nonlinearity The deviation between the (target position) and the actual coordinate position can be minimized over the entire surface of the substrate.

[0008]

In a semiconductor manufacturing process using a projection exposure apparatus, a first layer (fast exposure) pattern regularly arranged on a solid wafer using the projection exposure apparatus. The mask pattern of the second and subsequent layers is overprinted again using the projection exposure apparatus. That is, the first layer pattern formed by repeating the projection exposure by positioning the XY stage in a step-and-repeat manner in accordance with the arrangement coordinate position in the design is 2.
Exposure area (shot area) in exposure transfer after layer
In this exposure transfer, an alignment mark formed around the shot area as a part of the initial layer pattern is detected, and the mask pattern of the second and subsequent layers is exposed. The region is aligned two-dimensionally. Also, Japanese Unexamined Patent Publication No.
When performing step-and-repeat projection exposure using the method disclosed in Japanese Patent No. 44429, the alignment marks are detected only in appropriately selected 5 to 7 exposure areas on the wafer. Then, the coordinate positions of all the exposure regions on the wafer are calculated.

However, the initial layer pattern (exposure area) which is intended to be regularly arranged at the designed arrangement coordinate position using the projection exposure apparatus has an average of 0.05 μm on the actual wafer from the designed arrangement coordinate position. They tend to be formed with a certain degree of deviation, and the direction of each deviation is determined by the XY stage during exposure transfer.
It has been found that it is almost constant according to the moving direction of the step-and-repeat. A step-and-repeat path of a projection exposure apparatus for manufacturing a semiconductor usually scans (steps) the wafer in a zigzag manner, and the stepping movement in the opposite direction is performed between the even and odd rows of the initial layer pattern. Becomes At this time, the target stop position and the actual stop position of the XY stage are caused by mechanical factors such as slight backlash of the feed mechanism of the XY stage and variations in lubrication characteristics within an allowable range. It is considered that a uniform steady-state deviation amount corresponding to the direction of the stepping occurs between. Since this uniform steady-state deviation amount is usually lower than the minimum movable amount of the movable stage that can be controlled, it is easy to detect and correct the alignment mark again in each exposure area. However, the direction of the steady-state deviation of the actual array coordinate position of the first layer pattern exposed according to the designed array coordinate value is equal to each other between even-numbered rows or odd-numbered rows, and is opposite between even-numbered rows and odd-numbered rows. is there.

[0010] A steady-state deviation amount of the alignment in accordance with the step-and-repeat direction due to the mechanical factors of the projection exposure apparatus occurs every time the exposure transfer is performed even in the second and subsequent layers. Therefore, the second layer formed by repeating the projection exposure by stepping the XY stage in accordance with the coordinate position calculated by sample-aligning some of the initial layer patterns is the actual initial layer pattern. Includes the steady-state deviation amount with respect to the coordinate position (target position) of the exposure area calculated at the time of exposure of the second layer, in addition to the uniform steady-state deviation amount with respect to the array coordinate position in design.

Further, such step and repeat
The deviation amount of the regular alignment in accordance with the direction of the projection exposure apparatus strongly depends on the step-and-repeat direction, and also depends on the variation of mechanical factors between different projection exposure apparatuses. It has been confirmed that the distribution of the steady-state deviation amount on the wafer is well reproduced in the case of performing the overlap exposure using the same stepping path using the projection exposure apparatus.

The present invention utilizes this reproducible deviation amount distribution to calculate the target stop position of the movable stage reflecting the actual arrangement state of the area, or to calculate the target stop position. , The actual stop position of the movable stage is brought closer, and the positioning accuracy is increased to a level below the minimum controllable amount (corresponding to an error amount that cannot be absorbed by redoing) in the conventional movable stage. It is an object of the present invention to provide a positioning apparatus and a projection exposure apparatus.

[0013]

According to the first aspect of the present invention, a stage (WS) capable of mounting an object (W) on a position control device and moving in a two-dimensional plane, and Based on the feed error information (ΔMs, Δx1, Δx2) of the stage caused by the difference in the moving direction in the two-dimensional plane,
Control means for controlling the movement of the stage (B, N, HY,
CX, KY, MX), measurement means (E, F, G) for measuring the position of the stage in a two-dimensional plane, and position information of a plurality of regions on the substrate placed on the stage. And actual measurement means (KX) for performing actual measurement in cooperation with the control means, and the control means controls the movement of the stage based on the error information and the position information actually measured by the actual measurement means. According to the second aspect of the present invention, a substrate (W) in which a plurality of regions are regularly arranged in a two-dimensional manner based on design arrangement information is placed on the position control device, and the two-dimensional arrangement is performed. When a stage (WS) that can move in a two-dimensional plane along each of the directions and a plurality of regions linearly arranged in a predetermined arrangement direction on the substrate are formed as one set, the set (odd row ) And another set (even-numbered row) adjacent thereto, based on error information (ΔMs) from the designed arrangement state that can occur between the sets (even rows).
N, HY, CX, KY, MX) and measuring means (E, F, G) for measuring the position of the stage in a two-dimensional plane
And actual measurement means (KX) for actually measuring positional information of a plurality of regions on a substrate placed on a stage in cooperation with the measurement means.
And the control means controls the movement of the stage based on the error information and the position information actually measured by the actual measurement means. In the invention according to claim 4,
A substrate (W) in which a plurality of regions are regularly arranged in a two-dimensional manner based on design arrangement information is placed on the position control device, and a two-dimensional plane is arranged along each of the two-dimensional arrangement directions. A stage (WS) that can move in the inside, a measuring unit (KX) that measures the position of the stage, and positional information of a plurality of specific regions (sample shots) on the substrate are cooperated with the measuring unit. Actual measurement means (F, G) for actual measurement, storage means (N) for storing feed error information (ΔMs) which may occur when the stage is moved, position information of the actually measured specific area, and the design Control means (B, B) for controlling the movement of the stage based on the above arrangement information and the feed error information.
H, CX, MX). Further, in the invention according to claim 19, the stage (WS) on which an object (W) is placed and which can move in a two-dimensional plane is error information due to a difference in the moving direction in the plane and the stage (WS). Error information (ΔMs, Δx1, Δ
x2) is obtained, the position of the stage in the two-dimensional plane is measured, the position information of a plurality of regions on the substrate placed on the stage is actually measured in cooperation with the position measurement of the stage, and error information and The movement of the stage is controlled based on the actually measured position information. According to the twentieth aspect of the present invention, a substrate (W) in which a plurality of regions are regularly arranged in a two-dimensional manner based on design arrangement information is placed, and each of the substrates is arranged along each of the two-dimensional arrangement directions. In a position control method of a position control device (FIG. 1) having a stage (WS) movable in a two-dimensional plane, a plurality of regions linearly arranged in a predetermined arrangement direction on the substrate are defined as one set. Then, error information (ΔMs) from the designed arrangement state that may occur between the set and another adjacent set is obtained, the position of the stage in a two-dimensional plane is measured, and The position information of a plurality of regions on the substrate placed on is actually measured in cooperation with the position measurement of the stage, and the movement of the stage is controlled based on the error information and the actually measured position information. . In the invention according to claim 22, a substrate (W) in which a plurality of regions are regularly arranged two-dimensionally on the basis of arrangement information on a design is placed, and the substrate is placed in each of the two-dimensional arrangement directions. A position control method for a position control device (FIG. 1) having a stage (WS) movable along a two-dimensional plane along a plane, wherein the position of the stage is measured and position information of a plurality of specific regions on the substrate is obtained. Is measured in cooperation with the position measurement, and feed error information (ΔMs) that may be generated when the stage is moved is stored, and the position information of the actually measured specific region and a plurality of regions on the substrate are stored. The movement of the stage is controlled based on the array information in design and the feed error information.

[0014]

[0015]

[0016]

[0017]

According to the present invention, the movement of the stage is controlled in consideration of the feed error information resulting from the difference in the movement direction of the stage and the actual measurement result of the stage position. Thus, accurate positioning of the stage can be performed regardless of the moving direction of the stage. According to the second and twentieth aspects of the present invention, error information generated between adjacent rows of a plurality of regions on a substrate regularly arranged two-dimensionally, and the actual measurement result of the stage position are taken into consideration. Since the movement is controlled, accurate stage positioning can be performed irrespective of differences in rows in a plurality of two-dimensionally arranged areas. According to the inventions described in claims 4 and 22, when positioning a plurality of areas calculated by a statistical method, positioning is performed in consideration of feed error information generated due to movement of the stage. Even if a deviation from the target position depending on the moving direction of the stage occurs, the substrate can be accurately positioned at the target position.

In this embodiment, first, the coordinate position of the sampling shot area on the substrate is actually measured while moving the movable stage, and the actual measured value of the coordinate position of the sampling shot area and the array coordinate position in design are measured. After calculating the actual coordinate position of the shot area based on the above, the movable stage is moved, the movable stage is moved toward the calculated coordinate position of the positioning target position, and the movable stage is stopped. The alignment of the shot area is completed. Here, it is predicted that the coordinate position at which the movable stage actually stops at the time of completion of the alignment is regularly shifted from the coordinate position of the initial target position in a predetermined arrangement direction on the substrate. It is assumed that the amount and direction of the displacement within one linearly arranged set are substantially common to each other. Therefore, in order to improve the alignment accuracy by utilizing the regularity of the distribution of displacement on the substrate, the expected value of the relative placement error amount in each set is stored in a memory in advance, and the movable stage Is guided toward the coordinate position of the positioning target position and is stopped, the target position is corrected by the expected value of the arrangement error amount. That is, the set to which the positioning target position belongs is determined, the corresponding placement error amount is called from the memory, the positioning target position is corrected, and the corrected new positioning target position is indicated to the stage control system. The stage control system guides and stops the movable stage toward the corrected coordinate position.

According to this embodiment, first, the movable stay
After moving the image, the coordinate position of the sampling shot area on the substrate is actually measured, and the coordinate position of the shot area is calculated based on the actually measured value of the coordinate position of the sampling shot area and the designed array coordinate position. Stay
The movable stage is moved by the stage control system, and the movable stage is guided toward the calculated coordinate position of the positioning target position and stopped, thereby completing the positioning of the shot area. Here, the actual coordinate position of each shot area is deviated from the designed arrangement coordinate position, and the distribution tendency of the deviation on the substrate is such that a single row of shot areas linearly arranged in a specific arrangement direction on the substrate is used. When a set is used, some of the sets have almost the same arrangement error directionality.

Therefore, in order to improve the alignment accuracy by utilizing the tendency of the distribution of the shift on the substrate, at the stage of actually measuring the coordinate position of the sampling shot area, a set of sets having substantially the same directionality of the arrangement error is used. Only the inside sampling shot area is selected and measured. That is, a group that can be regarded as having substantially the same orientation of the placement error is previously selected, and when selecting a sampling region, only the region included in the selected group is specified.

According to this embodiment, first, the movable stay
After measuring the coordinate position of the sampling area on the substrate while moving the substrate, and calculating the coordinate position of the area based on the actual measurement value of the coordinate position of the sampling area and the array coordinate position in design, the stage control is performed. Movable stay depending on the system
Then, the movable stage is guided toward the coordinate position of the positioning target position to stop the movable stage, thereby completing the alignment of the area. Here, the actual coordinate position of each area is deviated from the designed arrangement coordinate position, and the distribution tendency of the deviation on the substrate is a plurality of regions consisting of one line of regions arranged linearly in a specific arrangement direction on the substrate. It is considered that the relative placement error amount is constant between at least two of the sets.

Therefore, in order to improve the positioning accuracy by utilizing the tendency of the distribution of the displacement on the substrate, when calculating the coordinate position of the area, the measured value of the coordinate position of the sampling area is calculated by the relative arrangement error amount. By applying a reverse bias, the coordinate position without the placement error is calculated. That is, the relative placement error amount occurring between at least two sets is stored in a memory in advance, and the correction unit stores the relative placement error amount corresponding to at least two sets including the sampling area. Then, the correction value is obtained by subtracting this relative placement error amount from each measured value. The calculating means calculates the coordinate position of the positioning target position based on the designed array coordinate position and the correction value of the coordinate position of the sampling area.

[0023]

[0024]

[0025]

An embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic view of a projection exposure apparatus according to an embodiment. This is a projection exposure apparatus for performing a step-and-repeat type projection exposure using the method introduced in Japanese Patent Application Laid-Open No. 61-44429, in which only a drive system in the X direction is shown for a wafer stage. , Y-direction drive systems are not shown.

In FIG. 1, an XY stage WS on which a wafer W in which a large number of exposure shot areas are regularly two-dimensionally arranged is mounted on a drive motor controlled by a stage drive control system CX. MX can move in the X direction parallel to the plane of the paper, and can cooperate with a similar drive system in the Y direction to send an arbitrary exposure area on the wafer W into the field of view of the projection optical system PL. The laser interferometer KX measures the coordinate position in the X direction of the XY stage WS in real time. The alignment shot specifying unit selects the wafer W from the combination pattern of the shot area distribution of the sample alignment stored in the storage unit M and the alignment path.
Optimal pattern according to the arrangement and number of shot areas above
One of the positions is selected, and each position of the selected sample alignment shot area is instructed to the stage drive control system CX. The stage drive control system CX drives the XY stage WS so as to scan the wafer W along the selected alignment path.

The actual measurement of the coordinate position in the selected sample alignment shot area (hereinafter referred to as the sampling area) is performed prior to the step-and-repeat type projection exposure, and the XY stage WS is moved. In some sample alignment shot areas on the wafer W, marks WM formed at corners of the shot areas
Is detected, and the coordinate position of the exposure shot area is defined in association with the coordinate position of the XY stage WS measured by the laser interferometer KX. The exposure position determination unit B
Statistical calculation is performed based on the actually measured coordinate position in the sampling area and the designed array coordinate value of the exposure shot area to determine the stop target coordinate position of the XY stage WS for all the exposure shot areas. calculate. The storage unit N
The expected amount and direction of a deviation that can occur between the target stop position and the actual stop position in each exposure area are stored in advance in association with the step-and-repeat path. The data is instructed to the correction unit H. The correction unit H performs an XY stage corresponding to each exposure area calculated by the exposure position determination unit B according to the expected amount and direction of the deviation (hereinafter, referred to as expected distribution).
The target stop position of the WS is corrected. Therefore, when the stage drive control system CX steps the XY stage according to the designated step-and-repeat path, the X-Y is measured every moment by the laser interferometer KX.
The XY stage is moved and stopped one after another so that the coordinate position of the Y stage coincides with the corrected target coordinate position.

As an alignment apparatus for aligning the mask pattern PA and the exposure shot area on the wafer W in a conjugate plane sandwiching the projection lens PL, an alignment mark WM in the exposure shot area is aligned with the projection lens PL and Mask R
, The TTR system E that can confirm the direct overlap between the alignment marks of the mask pattern P and the wafer W, and the mark WM through only the projection lens PL to detect the projection lens. Marker in the conjugate plane on the wafer side of PL
The mark WM is detected via a TTL system capable of confirming the position of the mark WM and a dedicated optical system fixed outside the projection lens PL, and the mark WM (and the shot area) for the projection lens PL is indirectly detected. The three systems of the off-axis system which can confirm the relative position of are shown.

In the projection exposure apparatus configured as described above, the actual measurement of the coordinate position of the sampling area by the alignment shot designating section C and the calculation and correction of each coordinate position of the exposure shot area by the exposure position determining section B are performed. Later, the step-and-repeat mechanical forward movement by the stage drive control system CX and the projection exposure at each stop position are repeated. At this time, based on the information stored in the storage unit M, if only sampling areas having the same arrangement error direction are selected, for example, only a shot area having the same stepping direction is selected, as compared with a case where random selection is performed. Therefore, the influence of the arrangement error at the calculated target coordinate position of the exposure shot area is small, and
Based on the estimated amount stored in the storage unit N, the correction unit H
Since the target stop position of the Y stage WS is corrected, the influence of the mechanical factors of the projection exposure apparatus on the final positioning accuracy is smaller than when the correction is not performed.

FIG. 2 is a plan view of the wafer, and FIG. 3 is a partially enlarged view of FIG. Here, for the initial layer pattern formed using a typical step-and-repeat path, the distribution state of the deviation amount between the designed arrangement coordinate position and the actual arrangement coordinate position is shown by this step-and-repeat. It is shown over the route.

In FIG. 2, a total of 32 exposure areas S1 to S32 are arranged on the wafer W. Exposure area S
In steps 1 to S32, a first layer pattern is projected and exposed on a solid wafer along a zigzag step-and-repeat path indicated by a dotted line by using a projection exposure apparatus, with a designed array coordinate position as a target position. It was formed as follows. Here, for each of the exposure areas S1 to S32 of the wafer W, the same projection exposure apparatus is used and the second layer mask pattern is formed along the same step and repeat path.
Are stacked in turn.

At this time, six of the exposure areas S1 to S32 are selected as sampling shot areas, the alignment marks are actually detected for the six exposure areas, and the coordinate positions of the exposure areas are measured. The coordinate positions of all the exposure areas S1 to S32 are calculated based on the six actually measured values and the design coordinate positions. Thereafter, the same step-and-repeat path as when the first layer pattern is projected and exposed is traced.
XY stages are successively guided toward the calculated coordinate positions of the exposure areas S1 to S32 while being stopped, and stopped.
The projection exposure of the layer is repeated.

The exposure regions S1 to S32 on the wafer W are obtained by projecting and exposing the initial layer pattern along the step-and-repeat path shown in FIG. 2, so that the step-and-repeat of the XY stage is performed. Among the operations, even rows (even arrays) and odd rows (odd arrays) having the same moving direction have a common directionality in a deviation amount from a design array coordinate position with respect to an actual coordinate position. , The first row, 3
Shift in the positive direction of the X-axis by a deviation amount [Delta] x ₁ in row, the second row is shifted in the negative direction of the X-axis by a deviation amount [Delta] x _2. In this case, the deviation amount Δx ₂ is larger than the deviation amount Δx ₁ due to the influence of the backlash of the XY stage.

Accordingly, from among the exposure areas S1 to S32,
The exposure areas S4, S16, S27, S24, S11, and S1 belonging to the odd-numbered rows that are not affected by the backlash are selected as the sampling areas そ れ ぞ れ, and the coordinate positions are actually measured. Based on the position, first, the coordinate positions of all the exposure areas in the odd-numbered rows are calculated by a statistical method, and then, the even numbers shifted in the negative direction from the exposure areas in the odd-numbered rows by the deviation amount (Δx ₂ + Δx ₁ ). The coordinate positions of all the exposure areas in the row are calculated. As a result, compared with the case where the sampling area is randomly selected from the exposure areas S1 to S32, the regular deviation Δx
₁ , the accuracy of position detection corresponding to Δx ₂ can be improved.

Next, when the XY stage is positioned in the step-and-repeat method as it is at the coordinate positions of the exposure areas S 1 to S 32 calculated exactly as described above, and the second layer of projection exposure is performed, the exposure area of the odd-numbered rows is obtained. In the position shifted by the deviation amount Δx _{1 in the} positive direction from the calculated coordinate position, the deviation amount Δ in the negative direction from the calculated coordinate position in the exposure area of the even-numbered row.
is predicted to exposure projection is performed by x ₂ shifted position.

Therefore, the target stop coordinate position of the XY stage is corrected to a coordinate position shifted by the predicted deviation amount. That is, in the exposure area of the odd-numbered row, the deviation amount Δx _{1 is} shifted in the negative direction from the calculated coordinate position, and in the exposure area of the even-numbered row, the deviation amount Δx is shifted in the positive direction from the calculated coordinate position.
The target stop coordinate position of the XY stage is changed to a position shifted by _two .

Incidentally, it is also possible to randomly select six sampling areas from the exposure areas S1 to S32. In this case, for the sampling region selected from the odd-numbered rows, the actual measurement value is corrected to a coordinate position shifted by the deviation amount Δx ₁ in the negative direction, and for the sampling region selected from the even-numbered lines, the actual measurement value is deviated in the positive direction. Quantity Δx ₂
By correcting the shifted coordinate positions, the actual array position can be calculated for each of the exposure regions S1 to S32 with reference to the lattice-like designed array coordinate position indicated by the dotted line in FIG. With respect to the array coordinate position in the design, the calculated coordinate position is shifted in the positive direction by the deviation amount Δx ₁ for the exposure area in the odd row, and the calculated coordinate position is shifted in the negative direction for the even row exposure area. The actual array coordinate position is calculated as the coordinate position shifted by the deviation amount Δx ₂ from the above.

Although the embodiment of the present invention has been described above, the above-described matters are schematically shown in flowcharts in FIGS. 4 to 7, and the sequence will be briefly described again based on them. FIG. 4 shows a main routine processed by the computer of the stepper shown in FIG. 1, in which units A, B, C, F (or G), H, M, N, and CX in FIG. 1 are used. Shall be.

First, in step 100, a positioning (feed) error amount ΔMs due to a difference in the stepping direction of the XY stage WS of the stepper is set. This error amount ΔMs may be treated separately when the stepping direction is positive (even line in FIG. 2) and negative (odd line in FIG. 2). It may be handled as a relative value of the feed position in the negative direction based on the feed position at the time.

In the next step 101, it is determined whether the wafer to be exposed is the first layer (1st) or the overprint (2nd). Step 1 for first layer grilling
In 02, the design coordinate values corresponding to the shot area to be exposed are read from the memory, and in step 103, it is determined whether the shot area belongs to an even-numbered row or an odd-numbered row. A value obtained by correcting the coordinate value by an error amount ΔMs is calculated as a target position.

In step 104, the XY stage WS is stepped to the target position.
At 05, the first layer pattern in the shot area is exposed. In the next step 106, it is determined whether or not all the shots have been exposed on the wafer. If the exposure of all the shots has been completed, the step-and-repeat exposure on one wafer is completed. If it is determined in step 106 that there are remaining shots, then in the next step 107, the first layer (1st baking) is performed.
Print). In this case, since the printing is the first printing, the sequence returns to step 102 and the exposure operation for the next shot is started.

On the other hand, in step 101, overprinting (2nd
If the “Print” task is selected, go to Step 1
At 08, data input is performed. This data entry is
Of either step 109 or step 110
In each case, as shown in FIG. 3 above.
As shown in FIG.
領域 x₁ , △ x_Two As data
Set. However, step 109 is the second print
The error tendency of the shot arrangement on the wafer for
Is an operation that simply reads it out,
Step 110 is a stepper TTR alignment system
E or operation to measure by TTL alignment system F
It is. As an example of the actual measurement method, the y-direction
Multiple x-direction marks in a shot area aligned in a row
The position is measured sequentially and each mark in the shot area on the even line
Average value of position X_evAnd each marker in the odd-numbered shot area
Average value X_odAnd find the difference between them
誤差 x₁₂And memorize it.
In the case of actual measurement, as shown in FIG.
Error in direction △ x₁ And the error amount in the negative direction △ x_Two And separately
It is difficult to find, so their difference, that is, even line
Relative amount of error Δx in the X direction between a row and an odd row ₁₂As
To handle. On the other hand, in step 109, the first pre
Used for printing the front layer by printing (or 2nd print)
Data of the stepper-specific feed error △ Ms
Even lines by transferring from the stepper
Error from the ideal value occurring in each of the odd-numbered rows
△ x₁ , △ x_Two Can be handled separately.

Next, in step 111, an alignment mode is selected. Here, the field-by-field alignment (F
FA) and enhancement global alignment (EGA). The FFA method is a step-and-repeat exposure method in which, similarly to die-by-die alignment, exposure and alignment of one shot area on a wafer are repeated for each shot area. However, the point different from the die-by-die method is that the position of the XY stage WS for aligning the shot area to be exposed with the reticle pattern projected image and the shot area by the alignment sensor (particularly, TTL alignment system F) are used. This means that the position of the XY stage WS for detecting the accompanying mark is shifted in a specific direction by a substantially constant amount. For this reason, after detecting a mark in a certain shot area, the XY stage is always fed by a fixed amount (about several hundred μm) based on the detected position and stopped, and then exposure is performed. For this reason, in the FFA method, only the feed error amount ΔMs of the XY stage WS deteriorates the positioning accuracy of the pattern image to be overlapped and exposed. In other words, the effects of the placement errors Δx ₁ and Δx ₂ of the shot area on the wafer can be theoretically ignored since the alignment is performed for each shot.

If the FFA system is selected in step 111, the design coordinate values of the shot area to be overlaid on the wafer are read out in step 112, and the XY stage WS reads it out in step 113. Stepping movement is performed as a target position. Thereafter, in step 114, a mark is detected for the shot area, and the correct alignment position of the shot area, that is, the feeding position of the XY stage is actually measured. Further, in step 115, the actually measured feed position is corrected in accordance with a preset feed error amount ΔMs. Here, if the magnitude of the error amount ΔMs is different from the magnitude of the error amount ΔMs in accordance with the feeding direction with respect to the shot area, the actually measured feeding position is corrected in the same manner as the correction of the difference in the stepping direction due to the sign (even and odd rows). Is set to a new target position by correcting the value of X−
Position the Y stage.

Next, in step 105, the exposure of the shot area is performed, and the determination of the end of all the shots in step 106 and the determination of the first print in step 107 are performed. Since the task is a 2nd print (overprinting) task, the sequence returns to step 112. As described above, the 2nd printing by the FFA method is repeated until the sequence of steps 112 to 115 and 105 to 107 is completed until the exposure for the last shot area is completed.

If the EGA system is selected in step 111, one of several EGA modes is selected in step 116. Here, four modes M1, M2, M3, and M4 are prepared in advance, and the operator presets which of the four modes to use. Therefore, in step 116, the sequence for automatically executing one mode set in advance proceeds. The four modes will be described in detail with reference to FIGS. 5 to 7 below. In any of the modes, the XY stage for all the shot areas to be exposed on the wafer is subjected to the statistical calculation processing by the EGA method. A final positioning target position of the WS is calculated. Therefore, the next step 1
In step 21, the XY stage is stepped on the basis of the target position. In step 122, the shot area is exposed. In step 123, it is determined whether or not all shots have been exposed. At that time, the operation from step 121 is repeated again.

Now, the four modes M1 and M of the EGA system will be described.
2, M3 and M4 are steps 117 and 1 in FIG.
18, 119, and 120. Among them, a part of the mode M2 of step 118 is shared with the mode M1 of step 117 as shown in FIG. In the mode M1 in FIG. 5, first, at step 130, design coordinate positions of N sample alignment shot areas are designated. As in the conventional case, the designation of the sample shot here designates a shot area located at each vertex of a regular polygon (preferably a pentagon or more) centered on the shot located at the center of the wafer. This mode M1 is selected especially on the assumption that the shot areas already formed on the wafer to be overprinted are arranged so as to correct the occurrence of the placement errors Δx ₁ and Δx ₂ due to the stepping direction. Is what it is. That is, at the time of the step-and-repeat exposure of the first layer (or the previous layer), the task of the first print in FIG. 4 (or the task such as the second print of the FFA method).
Is a preferred mode when performing overprinting on a wafer on which the above has been performed. Therefore, in this case, since the error amounts Δx ₁ and Δx ₂ as shown in FIG. 3 have been corrected to be negligible, in step 130 of FIG. There is practically no problem if is selected for sample alignment.

Next, in step 131, each mark of the designated sample shot area is detected by the TTL alignment system F, and the actually measured coordinate position of the sample shot area is measured. Then, it is determined in step 132 whether or not the sample alignment has been performed for all of the specified N shot regions. If not, the process returns to step 131. When the N actual measured values are obtained in this way, the target coordinate position of the XY stage WS for positioning each of all the shot areas on the wafer with respect to the projected image of the reticle pattern is calculated in step 133 by statistical calculation processing. Is determined by The sequence of steps 130 to 133 described above is the same as the technique disclosed in the above-mentioned Japanese Patent Application Laid-Open No. 61-44429.

However, in this embodiment, step 13
In the next step 134, the target position calculated in step 3 is corrected by the feed error amount ΔMs of the XY stage of the stepper. The correction calculation in step 134 has the same meaning as in steps 103 and 115 in FIG. Thus, the process of the mode M1 of the EGA system (step 117 in FIG. 4) is completed, and the process proceeds to step 121 in FIG.

On the other hand, the mode M of the EGA system shown in FIG.
2 is the error amount due to the even and odd arrangement of the shot areas on the wafer.
It is assumed that x ₁ and 場合 x ₂ occur to the extent that they cannot be ignored.
6, 137 are steps 130 in mode M1,
131 and 132 are exactly the same. The difference is that N
In step 138 after the actual measurement coordinate positions are measured,
It is determined whether each sample shot area belongs to an even-numbered row or an odd-numbered row, and the measured coordinate position is set to the error amount △ x ₁ set in step 109 or 110, depending on which one the sample shot area belongs to,
△ is to correct in x _2. For example, for the measured coordinate position of the sample shot area belonging to the odd-numbered row, FIG.
For definition, performs calculations as corrected by △ x ₁ in the negative direction. Incidentally, when not only be determined by the relative error amount △ X ₁₂ may be added as a correction value for positive and negative values of the half with respect to the measured position of each shot with odd rows and even rows.

The data of the actually measured position corrected as described above is processed in step 133 in the latter half of the mode M1, and after correcting the target position calculated in step 134 by the error amount ΔMs, Go to 121.

The mode M3 of the EGA system is processed according to the sequence shown in FIG. This mode M3 also has an arrangement error Δx between even rows and odd rows in the shot arrangement on the wafer.
₁ , △ x ₂ (or △ x ₁₂ ). First, in step 140 in FIG. 6, it is selected whether to perform sample alignment on a shot area belonging to an odd-numbered row (ODD) or to perform sample alignment on a shot area belonging to an even-numbered row (EVEN). This selection is set in advance by the operator, and in principle, the same effect can be obtained by selecting either one. However, pilot processing may be performed in advance to select the one with the higher effect.

In step 141, among the plurality of shot areas belonging to the odd-numbered rows, N pieces which are radially located at substantially equal distances from the center of the wafer are selected as sample shot areas, and their design coordinate positions are designated. Similarly, in step 142, N sample shot areas are selected from a plurality of shot areas belonging to the even-numbered rows, and design coordinate positions are designated.

The next steps 143, 144 and 145 are as follows:
Steps 131 (136) and 132 (13)
7) The processing is exactly the same as that of 133, and the positioning target position of the XY stage WS with respect to all the shot areas on the wafer is determined based on the measured N actual measured coordinate positions. The next step 146 checks whether the sample shot area belongs to an odd-numbered row or an even-numbered row. If the odd-numbered row (ODD) has been selected in the previous step 140, the determination in the step 146 is NO. , The next step 147 is executed, and if an even-numbered row (EVEN) is selected in step 140, step 148 is executed. In step 147, of the target positions for all the shot regions on the wafer, the target position for the shot region belonging to the even-numbered row is corrected by the difference between the error amounts Δx ₁ and Δx ₂ or the relative error amount Δx _12. . In step 148, the target position corresponding to the shot area belonging to the odd-numbered row is corrected by the difference between the error amounts Δx ₁ and Δx ₂ or the relative error amount Δx ₁₂ . This step 147, 1
The correction at 48 is necessary because the Cartesian coordinate point of the target position determined by the statistical operation at step 145 has been determined according to the sample shots belonging to only one of the odd rows and the even rows. Then, in the last step 149, the corrected target position of each shot area is further corrected and calculated with the XY stage feed error amount ΔMs. The processing in step 149 is the same as that in step 10
It has the same meaning as 3, 115 and 134.

FIG. 7 shows the sequence of the mode M4 of the EGA system. In this mode, the target positions of the shot area belonging to the odd-numbered row and the shot area belonging to the even-numbered row on the wafer are determined separately. First, in step 150, design coordinate positions of N sample shot areas belonging to even-numbered rows and N sample shot areas belonging to odd-numbered rows are designated. Therefore, this step 1
Step 50 is the same as executing steps 141 and 142 in FIG. 6 at the same time. Next, in step 151, sample alignment is executed. In step 152, it is checked whether or not the actually measured coordinate positions have been obtained for each of the 2N shot areas. These steps 151 and 152
Is substantially the same as the processing in the previous modes M1, M2, and M3. However, in order to increase the accuracy of the EGA method, the total number 2N of the sample shot areas is determined by the mode M1.
It is conceivable that the maximum value will be twice as large as that of the case of ~ M3, and a slight decrease in the throughput accompanying this is unavoidable.

In the next step 153, statistical calculation processing is performed using the actually measured coordinates of the N sample shot areas belonging to the even-numbered rows, and XY corresponding to all shot areas belonging to the even-numbered rows. The positioning target position of the stage is determined. In the next step 154, a statistical calculation process is performed using the actually measured coordinate values of the N sample shot areas belonging to the odd rows, and the target positions corresponding to each of all the shot areas belonging to the odd rows are determined.

In the next step 155, it is checked whether or not the conversion parameters obtained by the statistical calculation are to be corrected.
Then, the target position corresponding to each of all the shot areas on the wafer is corrected by the XY stage feed error ΔMs, and the process proceeds to step 121. This step 1
56 is exactly the same as the previous steps 134 and 149. If the parameter needs to be corrected in step 155, it is determined in step 157 whether the parameter needs to be corrected. Then, if no correction is necessary in step 158 (O
If it is K, the process proceeds to step 156. If correction is necessary, the corrected value is introduced into the previous steps 153 and 154, and the target position of each shot area is recalculated based on the corrected parameters. I do. Normally, one parameter correction is sufficient, and the next time step 155 is executed, the process automatically proceeds to step 156.

The parameter correction in step 157 is as follows.
This is performed for an inconvenient parameter if there is an inherent difference. That is, when the process is executed based on steps 153 and 154, two sets of conversion parameters are created. The conversion parameters are A, B in the following equations (1) and (2), where the design coordinate value of the shot area is (D _xn , D _yn ) and the target coordinate position is (F _xn , F _yn ). , C, D, O
_x, represented by O _y. F _xn = A · D _xn + B · D _yn + O _x (1) F _yn = C · D _xn + D · D _yn + O _y (2)

The two equations (1) and (2) are used in step 15
3, 154, but in the X direction, y
The offset amounts O _y and O _{x in the} directions are the placement error amounts △ x ₁ and △
with the difference a matter of course for x _2. However, the remaining four parameters A, B, C, and D are x, y of the wafer.
Since it is a value determined by the amount of expansion and contraction in the direction (γ _x , γ _y ), orthogonality (ω), and the residual rotation error (θ) of the shot array coordinates on the wafer with respect to the running coordinate system of the XY stage, And the value determined by the sample shots of the odd-numbered rows should originally have no difference. Therefore, in step 157, the differences between these four parameters A, B, C, and D are compared, and
If the difference is larger than within a certain range, a value obtained by interpolating between the two parameters is determined as a correction parameter value. The transformation parameters determined here at step 153 and _{A 1, B 1, C 1} , D 1, the transformation parameters determined in step 154 as _{A 2, B 2, C 2} , D 2, A 1 and A _When the difference between the _two is large, for example, (A ₁ + A
₂ ) / 2 is used as a correction value, and is fed back to the target position calculation process (the calculation process of equations (1) and (2)) in steps 153 and 154 again. The amounts of expansion and contraction (γ _x , γ _y ), orthogonality (ω), and residual rotation (θ) are defined as follows. γ _x = A, γ _y = D, θ = C / D, ω = −B / A−C /
D

As described above, according to the sequence algorithm shown in FIG. 4 to FIG.
In any case of d printing, the mechanical alignment of the XY stage is corrected to achieve accurate alignment.

[0062]

According to the present invention, the movement of the stage is controlled in consideration of the feed error information caused by the difference in the moving direction of the stage and the actual measurement result of the stage position. Therefore, accurate stage positioning can be performed regardless of the moving direction of the stage. According to the second and twentieth aspects of the present invention, error information generated between adjacent rows of a plurality of regions on a substrate regularly arranged two-dimensionally, and the actual measurement result of the stage position are taken into consideration. Since the movement is controlled, accurate stage positioning can be performed irrespective of differences in rows in a plurality of two-dimensionally arranged areas. According to the inventions described in claims 4 and 22, when positioning a plurality of areas calculated by a statistical method, positioning is performed in consideration of feed error information generated due to movement of the stage. Even if a deviation from the target position depending on the moving direction of the stage occurs, the substrate can be accurately positioned at the target position. In this embodiment, since the error information to be corrected is a deviation amount having a reproducibility that is predicted in advance, a high positioning accuracy is realized by performing the stage drive control by subtracting this deviation amount from the drive target position. it can. Further, in this embodiment, since the coordinate position is actually measured only in the sampling region having the same orientation of the placement error, the coordinate position of all the regions is calculated with high accuracy by subtracting the deviation amount of the same tendency of each region on the substrate. it can. Further, according to the present embodiment, since the actually measured value of the coordinate position of the sampling area is corrected by the deviation amount of the code according to the directionality of the placement error, each of the positions on the substrate can be obtained even from the sampling areas having different placement error directions. The coordinate positions of all the regions can be calculated with high accuracy by subtracting the regular deviation amount of the regions.

[0063]

[0064]

[0065]

[Brief description of the drawings]

FIG. 1 is a schematic view of a projection exposure apparatus according to an embodiment.

FIG. 2 is a plan view of a wafer.

FIG. 3 is a partially enlarged view of FIG. 2;

FIG. 4 is a flowchart for explaining the sequence of the embodiment.

FIG. 5 is a flowchart for explaining the sequence of the embodiment.

FIG. 6 is a flowchart for explaining the sequence of the embodiment.

FIG. 7 is a flowchart for explaining the sequence of the embodiment.

[Explanation of symbols]

 A Alignment processing unit B Exposure position determination unit C Alignment shot designation unit E TTR system F TTL system G Off-axis system M Storage unit N Storage unit H Correction unit R Mask W Wafer AX Optical axis CX Stage drive control system KX Interference system MX Drive motor PL Projection optical system PA Mask pattern WM Positioning mark WS XY stage

 ────────────────────────────────────────────────── ─── Continuation of front page (54) [Title of the Invention] Position control device, exposure apparatus having the position control device, device manufactured by the exposure device, position control method, and position control method Exposure method using the same, and device manufacturing method including the exposure method

Claims (26)

(57) [Claims] 1. A stage on which an object is placed and which can move in a two-dimensional plane, based on feed error information of the stage caused by a difference in a moving direction of the stage in the two-dimensional plane. Control means for controlling the movement of the stage, measuring means for measuring the position of the stage in the two-dimensional plane, position information of a plurality of regions on the substrate mounted on the stage, the measuring means Actual measurement means for collaboratively measuring,
The position control device, wherein the control means controls the movement of the stage based on the error information and the position information actually measured by the actual measurement means. 2. A substrate on which a plurality of regions are regularly arranged in a two-dimensional manner based on design arrangement information is placed, and can be moved in a two-dimensional plane along each of the two-dimensional arrangement directions. And a plurality of regions linearly arranged in a predetermined arrangement direction on the substrate as one set, the design arrangement state that may occur between the set and another adjacent set. Control means for correcting and controlling the movement of the stage based on the error information from; measuring means for measuring the position of the stage in the two-dimensional plane; and a plurality of substrates on the substrate mounted on the stage. Actual measurement means for actually measuring the position information of the area in cooperation with the measurement means;
The position control device, wherein the control means controls the movement of the stage based on the error information and the position information actually measured by the actual measurement means. 3. The actual measuring means actually measures position information of a plurality of specific areas among the plurality of areas on the substrate, and the control means measures the error information and the position information of the actually measured specific area. 3. The position control device according to claim 1, wherein the movement of the stage is controlled based on and design arrangement information of the plurality of regions on the substrate. 4. 4. A substrate on which a plurality of regions are regularly arranged in a two-dimensional manner based on design arrangement information is placed, and moves in a two-dimensional plane along each of the two-dimensional arrangement directions. Possible stage, measuring means for measuring the position of the stage, position information of a plurality of specific areas on the substrate, actual measuring means for actually measuring in cooperation with the measuring means, when moving the stage Storage means for storing information on a possible feed error; control means for controlling the movement of the stage based on the actually measured position information of the specific area, the layout information on the design, and the feed error information And a position control device comprising: 5. A first calculating means for calculating the position information of the plurality of areas by a statistical calculation based on the actually measured position information of the specific area and the designed arrangement information. Correction calculation means for calculating target position information on which the stage is to be moved by performing a correction calculation based on the error information with respect to the calculation result of the first calculation means; and the target position information and the measurement position of the measurement means. 5. The position control device according to claim 3, further comprising: a movement control unit configured to control the movement of the stage based on the control. 6. The method according to claim 1, wherein the position information of the specific area actually measured by the actual measurement means is generated on the substrate due to the stage moving direction and generated between a target stop position of the stage and an actual stop position. An actual measurement position correction unit that performs a correction operation based on the arrangement error amount, wherein the first arithmetic unit statistically calculates the position information based on a calculation result by the actual measurement position correction unit and the array information on the design. The position control device according to claim 5, wherein the position is calculated by an operation. 7. A plurality of arrangement error amounts are provided in accordance with a moving direction of the stage, and the actually measured position correction unit is configured to perform the measurement based on the arrangement error amounts individually corresponding to the plurality of specific regions. The position control device according to claim 6, wherein the position information of each specific area is corrected. 8. When a plurality of regions linearly arranged in a predetermined arrangement direction on the substrate are set as one set, a direction of a relative placement error generated between the plurality of sets is determined. Extracting means for extracting a set which can be regarded as substantially the same; selecting means for selecting the specific plurality of regions actually measured by the actual measuring means from the sets extracted by the extracting means; calculating by the first calculating means Among the obtained position information, the position information of an area belonging to the group not extracted by the extraction means is attributed to the stage moving direction and between the target stop position of the stage and the actual stop position. Non-extracted group correction means for performing a correction operation based on the generated placement error amount, wherein the second operation means performs the correction operation result with respect to the position information of the area corrected by the non-extracted group correction means, For other areas, 6. The position control device according to claim 5, wherein the target position information is obtained by performing a correction operation on the calculation result of the calculating means based on the error information. 9. The arrangement error amount includes an error generated when the stage moves in a first direction in the two-dimensional plane, and an error generated when the stage moves in a direction opposite to the first direction. The position control device according to any one of claims 6 to 8, further comprising a relative error amount with respect to the error to be performed. 10. The apparatus according to claim 1, wherein the arrangement error amount is a first error generated when the stage moves in a first direction in the two-dimensional plane.
The position according to any one of claims 6 to 8, comprising an error amount and a second error amount generated when moving in a second direction opposite to the first direction. Control device. 11. An average of positions of every other region in a direction orthogonal to the predetermined direction among a plurality of regions arranged linearly in a predetermined direction on the substrate, and a predetermined direction of a remaining region. The position control device according to any one of claims 6 to 9, further comprising: an error amount measurement unit that obtains a difference from an average of positions in the orthogonal direction as the arrangement error amount. 12. The position according to claim 6, further comprising an error amount storage unit that stores the arrangement error amount in advance in association with a moving direction of the stage. Control device. 13. When a plurality of regions linearly arranged in a predetermined arrangement direction on the substrate are set as one set, a direction of a relative arrangement error generated between the plurality of sets is determined. Separating means for separating into a first group, which is a group of sets in a first direction, and a second group, which is a group of groups in which the directionality of the placement error is opposite to the first direction; A first selection unit that selects a first specific plurality of regions actually measured by the actual measurement unit from a group; and a second specific plurality of regions actually measured by the actual measurement unit from the second group. And a second selecting unit that selects the first group based on the actually measured position information of the first specific region and the designed array information. A first part for calculating position information of a plurality of regions by statistical calculation, A second unit that calculates position information of a plurality of regions in the second group by a statistical operation based on the position information of the second specific region and the design sequence information; 6. The position control according to claim 5, wherein the second calculation unit obtains the target position information by performing a correction operation on the calculation results of the first unit and the second unit based on the error information. 7. apparatus. 14. A first conversion parameter obtained by a statistical operation of the first part for converting the design sequence information into position information of a plurality of regions in the first group, and If the second conversion parameter obtained by the statistical operation of the first part for converting the array information of the second group into the position information of a plurality of regions in the second group is in a predetermined relationship, the two parameters are converted. The apparatus further comprises correction means for correcting, wherein the first and second parts of the first calculation means recalculate position information of a plurality of regions in each group based on the corrected conversion parameters. The position control device according to claim 13, wherein: 15. A plurality of the error information are provided according to a moving direction of the stage, and the control means is configured to control the feed error amount according to a moving direction of the stage with respect to each of the plurality of regions. 15. The position control device according to claim 1, wherein the movement of the stage is controlled based on the position. 16. The apparatus according to claim 1, wherein said feed error information transferred from an exposure apparatus which has baked a previous layer formed on said substrate is stored.
The position control device according to any one of claims 1 to 4. 17. An exposure apparatus, wherein a substrate positioned by the position control device according to claim 1 is exposed with a mask pattern. 18. A device manufactured through a step of exposing a device pattern on the substrate using the exposure apparatus according to claim 17. 19. A stage on which an object is placed and which can move in a two-dimensional plane acquires feed error information resulting from a difference in a moving direction in the plane, and the position of the stage in the two-dimensional plane The position information of a plurality of areas on the substrate placed on the stage is measured in cooperation with the position measurement of the stage, based on the error information and the actually measured position information. And controlling the movement of the stage. 20. A substrate on which a plurality of regions are regularly arranged in a two-dimensional manner on the basis of arrangement information in a design is placed, and can be moved in a two-dimensional plane along each of the two-dimensional arrangement directions. A position control method for a position control device having a plurality of stages, wherein, when a plurality of regions linearly arranged in a predetermined arrangement direction on the substrate are formed as one set, the position of the set and the adjacent set are determined. Acquiring error information from the design arrangement state that may occur between the two, the position of the stage in the two-dimensional plane is measured, and position information of a plurality of regions on a substrate placed on the stage Is actually measured in cooperation with the position measurement of the stage, and the movement of the stage is controlled based on the error information and the actually measured position information. 21. In the actual measurement, position information of a specific plurality of regions among a plurality of regions on the substrate is actually measured, and the error information, the actually measured position information of the specific region, and 21. The position control method according to claim 19, wherein the movement of the stage is controlled based on design arrangement information of a plurality of regions. 22. A substrate on which a plurality of regions are regularly arranged two-dimensionally based on design arrangement information is placed, and moves in a two-dimensional plane along each of the two-dimensional arrangement directions. A position control method of a position control device having a possible stage, wherein the position of the stage is measured, and position information of a plurality of specific regions on the substrate is actually measured in cooperation with the position measurement. Storing the feed error information that can occur when moving the, the measured position information of the specific area, the array information of the design of the plurality of areas on the substrate, based on the feed error information, A position control method comprising controlling movement of a stage. 23. In the control of the movement of the stage, the position information of the plurality of regions is calculated by a statistical operation based on the actually measured position information of the specific region and the design sequence information. The calculation result of the position information of the plurality of areas is corrected and calculated based on the error information to obtain target position information to be moved by the stage, and the stage movement is performed based on the target position information and the measured position. 22. The method according to claim 21, wherein
Or the position control method according to claim 22. 24. The error information is provided as a plurality of feed error amounts of the stage according to the moving direction of the stage, and the feed error amount according to the moving direction of the stage with respect to each of the plurality of regions. The position control method according to any one of claims 19 to 23, wherein the movement of the stage is controlled based on the following. 25. An exposure method comprising exposing a substrate positioned by the position control method according to claim 19 to a mask pattern. 26. A device manufacturing method, comprising a step of exposing a device pattern on the substrate using the exposure method according to claim 25.