JPH10284396A - Method for alignment and method for measuring alignment precision - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a high precision, high throughput, in-shot multipoint alignment method. SOLUTION: Positions of first marks MA in extended graphics adapter(EGA) shots S2 , S4 , S5 , S16 , S17 , S28 , S29 , S31 and second marks MB in a range surrounded by circles A through D are measured respectively. Coordinates of each shot region S on a stationary coordinate system are then calculated by EGA operation using the measurements of first marks and in-shot correction parameters are calculated by a statistic processing similar to EGA using the measurements of second marks. Since the objects, i.e., the second marks MB, are arranged in the vicinity of adjacent parts in adjacent shot regions, movement of a substrate W is small even when the position of the second marks MB are measured individually and the measuring time can be shortened as compared with that of conventional in-shot multipoint alignment.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an alignment method and an overlay accuracy measuring method, and more particularly, to an exposure apparatus used for manufacturing a semiconductor device, a liquid crystal display device, and the like in a photolithography process. The present invention relates to a preferred alignment method and a method of measuring overlay accuracy performed for evaluating the performance of an exposure apparatus.

[0002]

2. Description of the Related Art In a photolithography process for manufacturing a micro device such as a semiconductor device, a liquid crystal display device, an image pickup device (CCD) or a thin film magnetic head, a photomask or a reticle (hereinafter, collectively referred to as a "reticle"). There is used a projection exposure apparatus that projects a pattern formed on a substrate (hereinafter, referred to as a “wafer”) such as a wafer or a glass plate coated with a photosensitive material (photoresist) via a projection optical system. .

In this type of projection exposure apparatus, it is necessary to perform high-precision alignment (alignment) between a reticle and a wafer prior to exposure. In order to perform this alignment, a position detection mark (alignment mark) formed (exposed and transferred) in the previous photolithography process is provided on the wafer. By detecting the position of the alignment mark, the wafer is detected. (Or the circuit pattern on the wafer) can be accurately detected.

An off-axis type alignment microscope (measurement microscope) used for detecting the position of an alignment mark on a wafer is disclosed, for example, in Japanese Patent Laid-Open No. 2-541.
One having an image capturing and image processing function as disclosed in Japanese Patent Application Laid-Open No. 03-003 is known.

[0005] In the alignment process using such an alignment microscope, for example, Japanese Patent Application Laid-Open No. 61-44429.
As described in Japanese Patent Application Laid-Open No. H11-260, the positions of several (usually several to about ten) alignment marks formed in advance on the wafer are determined by the image processing signals of the alignment microscope and the wafer. Detection is performed based on a measurement value of an interferometer that monitors the position of the stage, and a statistical operation such as a least squares method is performed using the detection result and design data of a shot array on a wafer to determine a position for exposure. So-called enhanced global alignment (hereinafter referred to as “EG”) for determining the coordinate position of each shot on the wafer for
A ") was mainly used.

In recent years, as the degree of integration of integrated circuits has increased, the alignment accuracy required for a projection exposure apparatus has become stricter. In particular, the displacement of an alignment mark caused by deformation of a wafer due to heat treatment or the like has occurred. In addition, misalignment of the alignment mark due to distortion (so-called distortion) of the projection optical system is also a hindrance to the improvement of the alignment accuracy. Alignment marks are prepared in advance, and a plurality of marks are measured even in the shot area during alignment measurement.
A method of obtaining not only the array coordinates of shots but also at least one of a rotation error and a shape error with respect to the reticle pattern in each shot area by the same statistical calculation as in A, and using it as a correction value at the time of position alignment (hereinafter referred to as "shot (Referred to as “inner multi-point alignment”).
No. 75496.

In the case of such an intra-shot multipoint alignment, the positions of a plurality of alignment marks arranged in an arbitrarily set alignment shot are individually measured, and the measurement result and the design of the reference position in each shot area are determined. By performing predetermined statistical processing using the array coordinates and the designed relative array coordinates of the respective alignment marks with respect to the reference position, the parameters for determining the array coordinates of the plurality of shot areas and the in-shot correction parameter (shot magnification) Correction parameters and shot rotation correction parameters).

FIG. 9 is a schematic diagram for explaining this multipoint alignment within a shot in a manner that is easy to understand visually. In FIG. 9, each rectangular area defined by a dotted line indicates a designed shot area, a rectangular area indicated by a solid line indicates an actual shot area, and a hatched shot area S is Shows an arbitrary alignment shot. Briefly, multi-point alignment within a shot is a method of measuring the positions of a plurality of (four in this case) alignment marks M _{S1 to} M _S4 formed on an alignment shot S, and as a result of the measurement. The position shift amounts indicated by arrows A1 to A4 in FIG. 9 are obtained from each of the alignment shots, and the position shift amounts are statistically processed. It is trying to get.

Further, similarly to the above, a plurality of measurement marks arranged in an arbitrarily set measurement shot are individually measured to calculate an in-shot correction parameter.
2. Description of the Related Art Registration measurement (vernier measurement) for evaluating the performance of an exposure apparatus by measuring the overlay accuracy of a plurality of circuit patterns formed by overlay exposure is known. The measurement of the overlay accuracy is performed using a registration measurement device. As the registration measurement mark, a box-in-box mark, which is a large and small square mark whose mark center is set at the same coordinate position by design, is used. Used. In order to more accurately measure the overlay accuracy, the normal registration measurement marks are arranged at the corners of each shot area, similarly to the alignment marks in FIG.

[0010]

As described above, in the intra-shot multipoint alignment, it is necessary to individually measure the positions of a plurality of alignment marks arranged in an arbitrarily set alignment shot. In order to improve the measurement accuracy due to the averaging effect, it is necessary to select a plurality of shot regions by performing an alignment shot. There is a disadvantage that the throughput is reduced.

That is, for example, when eight shots are selected as sample shots, in the case of EGA, when the alignment mark is a two-dimensional mark (a mark in which the XY coordinate position can be obtained from one mark), the number of measurements is made for each alignment shot. In the case of performing multi-point alignment within a shot, it is assumed that four alignment marks exist in one shot as shown in FIG. Four measurements are required for the sample shots, for a total of 8 × 4 = 32 measurements.

The positions of the marks in each shot area are determined by detecting a plurality of measurement errors in the same shot area in order to detect a rotation error and a shape error (magnification error) with respect to the reticle pattern in each shot area. Obviously, it is not preferable to arrange the marks so that all the marks are adjacent to each other. For example, when four alignment marks are arranged in the shot area, they may be arranged at the corners as shown in FIG. desirable.

Therefore, when sequentially measuring alignment marks in the same shot, it is necessary to move the wafer stage by a certain distance (a distance substantially equal to the length of the shot), which also increases the measurement time. Let
This was a factor that reduced the throughput.

Similarly, the problem that the number of times of measurement is large and the time required for measurement is enormous also exists in the registration measurement described above.

In any case, the number of measurements is reduced by reducing the number of marks in each shot area.
A decrease in the number of marks leads to a decrease in measurement accuracy, and is not a very realistic choice.

The present invention has been made under such circumstances, and an object of the present invention is to provide a positioning method capable of performing high-precision, high-throughput multipoint alignment in a shot. is there.

Another object of the present invention is to provide a positioning method capable of further improving the throughput, in addition to the object of the present invention.

Another object of the present invention is to provide a method of measuring overlay accuracy with high accuracy and high throughput.

[0019]

In the conventional multi-point intra-shot alignment, the shot arrangement is performed by a single statistical processing operation using the measurement results of a plurality of marks arranged in the same sample shot area in common. And an in-shot correction parameter are determined (see JP-A-6-275496). Therefore, if the number of sample shots and marks is increased in order to improve the measurement accuracy, the number of measurements will inevitably increase in accordance with the number, and the measurement time will increase due to the movement of the substrate when measuring each mark. It is thought to be. However, it is considered that the measurement result of the same mark does not always need to be commonly used for determining the parameters for the shot arrangement and for determining the intra-shot correction parameter. In order to accurately calculate the shot arrangement coordinates of all shots on the substrate, shot regions located at different positions on the substrate (a plurality of the shot regions are preferably peripheral shots) are selected as sample shots. Although it is necessary, the rotational deviation and the shape error of the pattern in each shot area occur in all shot areas in the same manner, so that the in-shot correction parameter can be obtained by using the measurement results of the marks in any shot area. Can be calculated accurately. The present invention has been made in view of these points, and employs the following solutions. The invention according to claim 1, wherein each of a plurality of shot areas (S) arranged on a substrate (W) and onto which a pattern on a mask (R) is transferred is a stationary coordinate system that defines a movement position of the substrate. In order to align the plurality of shot areas with each other on the stationary coordinate system, at least one of a rotation error and a shape error of each shot area with respect to the pattern on the mask in order to align the plurality of shot areas with each other. And a plurality of shot areas selected in advance among a plurality of position detection marks arranged in a fixed relative positional relationship with respect to a reference position in each of the shot areas. And a plurality of second marks arranged in the vicinity of the adjacent portions in the pre-selected mutually adjacent shot areas. A first step of measuring the position of the mark (MB); the arrangement result of the measurement of the first step and the reference position in each of the shot areas, and the first mark and the second mark with respect to the reference position. By performing predetermined statistical processing using the relative arrangement coordinates in the design of the mark, the arrangement coordinates of the plurality of shot areas on the stationary coordinate system and the rotation of each shot area with respect to the pattern on the mask are determined. A second step of obtaining at least one of an error and a shape error.

According to this, in the first step, the position of at least one first mark arranged in each of the plurality of shot regions selected in advance and the position of the first mark in the adjacent shot region selected in advance are determined. The positions of the plurality of second marks arranged in the vicinity of the section are respectively measured. Here, the plurality of shot areas where the first mark to be measured exists and the mutually adjacent shot areas where the second mark to be measured exists are not necessarily the same shot area. In addition, since the plurality of second marks to be measured are arranged in the vicinity of the adjacent portion in the previously selected mutually adjacent shot areas, when measuring these second marks, Even if the position of the second mark is measured individually, the movement of the substrate at the time of the measurement is slight, and the measurement time can be reduced as compared with the conventional measurement of the multipoint alignment within a shot.

In the second step, the measured result of the first step, the designed arrangement coordinates of the reference position in each shot area, and the designed relative arrangement coordinates of the first mark and the second mark with respect to the reference position are obtained. By performing a predetermined statistical process using each of them, an arrangement coordinate of a plurality of shot areas on a stationary coordinate system and at least one of a rotation error and a shape error of each shot area with respect to a pattern on a mask are obtained. That is, the calculation of the array coordinates of the plurality of shot areas on the stationary coordinate system and the calculation of at least one of the rotation error and the shape error (intra-shot correction parameter) with respect to the pattern on the mask in each shot area are:
Performed by separate statistical processing.

As described above, according to the first aspect of the present invention, the time required for the conventional EGA (measurement of one point in a shot) can be reduced to the vicinity of an adjacent portion in a preselected mutually adjacent shot area. Only a slight measurement time for measuring the positions of the plurality of second marks arranged at the position increases. Therefore, it is possible to perform high-throughput alignment with high accuracy and a short time required for measurement, as in the conventional multipoint alignment within a shot.

Of course, a part of the first mark to be measured and the second mark are
Part of the mark may be the same mark, and in such a case, the measurement time can be shortened accordingly.

In this case, when measuring the positions of the plurality of second marks arranged near the adjacent portions in the mutually adjacent shot areas in the first step, the substrate is moved little by little to measure the positions of the second marks. Two marks may be measured sequentially,
According to a second aspect of the present invention, the position measurement of the plurality of second marks in the mutually adjacent shot areas in the first step may be performed while the substrate is stationary.
By doing so, the measurement time is further shortened, and the measurement is performed in a stationary state of the substrate. For example, when the position of the substrate is measured by a laser interferometer or the like, a measurement error due to air fluctuation or the like is caused. Can be prevented from occurring.

Simultaneous measurement of the positions of a plurality of second marks arranged in the vicinity of an adjacent portion in a mutually adjacent shot area as described above is performed, for example, in such a case that the image field of the objective lens is wide and each image field is within the image field. This can be realized by using a low-magnification alignment microscope having a plurality of detection criteria corresponding to the second mark.

According to a third aspect of the present invention, a first measurement mark (M1) exposed simultaneously with a new pattern on a plurality of shot areas on a substrate, and an existing pattern formed on each of the shot areas earlier than the first measurement mark (M1). , A second measurement mark (M) whose design relative positional relationship with the first measurement mark is known.
2) measuring the relative positional relationship between the first measurement mark and the second measurement mark corresponding to each of the first measurement marks. A first step of measuring a relative positional relationship between a plurality of first measurement marks arranged in the vicinity of an adjacent portion in the inside and a second measurement mark corresponding thereto; predetermined statistics using the measurement results of the first step And a second step of performing processing.

According to this, in the first step, from among the plurality of first measurement marks and the second measurement marks corresponding to each of the first measurement marks, the plurality of first measurement marks are located in the vicinity of the adjacent portion in the mutually adjacent shot area. The relative positional relationship between the plurality of arranged first measurement marks and the corresponding second measurement marks is measured, and in the second step, predetermined statistical processing is performed using the measurement results, and the overlay accuracy is measured. Is done. Here, since the plurality of first measurement marks to be measured and the second measurement marks corresponding to the first measurement marks are arranged in the vicinity of the adjacent portions in the mutually adjacent shot areas which are selected in advance, these The movement of the substrate when measuring the relative positional relationship between the first mark and the second mark is small, and the measurement does not take much time. Therefore, it is possible to measure the overlay accuracy with high accuracy and high throughput.

[0028]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to FIG.
7 will be described with reference to FIG.

FIG. 1 shows a projection exposure apparatus 100 for implementing the alignment method according to the present invention.
The projection exposure apparatus 100 is a step-and-repeat type reduction projection exposure apparatus (a so-called stepper).

The projection exposure apparatus 100 includes an illumination system IOP
A reticle stage RST for holding a reticle R as a mask, a projection optical system PL for projecting an image of a pattern formed on the reticle R onto a wafer W as a substrate coated with a photosensitive material (photoresist), An XY stage 20 that holds a wafer W and moves on a two-dimensional plane (within an XY plane), a drive system 22 that drives the XY stage 20, and C
A main control device 2 comprising a minicomputer (or microcomputer) including a PU, a ROM, a RAM, an I / O interface, and the like, and integrally controlling the entire device.
8 is provided.

The illumination system IOP includes a light source (a mercury lamp or an excimer laser or the like) and an illumination optical system including a fly-eye lens, a relay lens, a condenser lens, and the like. The illumination system IOP illuminates a pattern on the lower surface (pattern forming surface) of the reticle R with a uniform illuminance distribution by exposure illumination light IL from a light source. here,
As the illumination light for exposure IL, a bright line such as an i-line of a mercury lamp, or an excimer laser beam such as KrF or ArF is used.

A reticle R is fixed on the reticle stage RST via fixing means (not shown). The reticle stage RST is driven by a driving system (not shown) in the X-axis direction (perpendicular to the plane of FIG. 1) and in the Y direction. Axial direction (left-right direction in FIG. 1) and θ direction (rotation direction in XY plane)
Can be finely driven. Thus, reticle stage RST can position reticle R (reticle alignment) in a state where the center of the pattern of reticle R (reticle center) substantially matches optical axis AXp of projection optical system PL. FIG. 1 shows a state in which this reticle alignment has been performed.

The projection optical system PL has its optical axis AXp in the Z-axis direction orthogonal to the plane of movement of the reticle stage RST. In this case, the projection optical system PL is telecentric on both sides and has a predetermined reduction magnification β (β is, for example, １ ／). Things are used. For this reason, if the reticle R is illuminated with uniform illuminance by the illumination light IL in a state where the pattern of the reticle R and the shot area on the wafer W are aligned (aligned), the pattern on the pattern forming surface is changed. Projection optical system P
L, the image is projected on the wafer W coated with the photoresist, and a reduced image of the pattern is formed in each shot area on the wafer W.

The XY stage 20 is actually a Y stage that moves on a base (not shown) in the Y axis direction.
An X stage is configured to move on the stage in the X axis direction. In FIG.
Shown as zero. A wafer table 18 is mounted on the XY stage 20, and a wafer W is held on the wafer table 18 by vacuum suction or the like via a wafer holder (not shown).

The wafer table 18 minutely drives the wafer holder for holding the wafer W in the Z-axis direction, and is also called a Z stage. This wafer table 18
A movable mirror 24 is provided on the upper surface of one end of the laser table, and a laser beam is projected onto the movable mirror 24 and the reflected light is received to measure the position of the wafer table 18 in the XY plane. An interferometer 26 is provided to face the reflecting surface of the movable mirror 24. Note that the moving mirror is actually X
An X moving mirror having a reflecting surface orthogonal to the axis and a Y moving mirror having a reflecting surface orthogonal to the Y axis are provided. In correspondence with this, the laser interferometer also has an X laser interferometer for measuring the position in the X direction. Although a Y laser interferometer for Y-direction position measurement is provided, these are representatively shown as a movable mirror 24 and a laser interferometer 26 in FIG. Therefore, in the following description, the XY of the wafer table 18 is controlled by the laser interferometer 26.
Assume that the coordinates are measured. In addition, the laser interferometer 26
Is referred to as a stage coordinate system in the following description.

The measured values of the laser interferometer 26 are stored in the main controller 2
The main controller 28 controls the laser interferometer 2
By driving the XY stage 20 via the drive system 22 while monitoring the measurement value of the
Is positioned. In addition, the output of a focus sensor (not shown) is also supplied to the main controller 28, and the main controller 28 controls the drive system 2 based on the output of the focus sensor.
Then, the wafer table 18 is driven in the Z-axis direction (focus direction) via the interface 2. That is, the wafer W is positioned in the three axes of X, Y, and Z via the wafer table 18 in this manner.

On the wafer table 18, a reference plate FP whose surface is the same as the surface of the wafer W is provided.
Has been fixed. On the surface of the reference plate FP, various reference marks including a reference mark used for so-called baseline measurement or the like are formed.

Further, in this embodiment, an off-axis type alignment sensor AS for detecting an alignment mark (mark for position detection) formed on the wafer W is provided on a side surface of the projection optical system PL. As the alignment sensor AS, an imaging type alignment sensor of an image processing system is used.

Next, a specific configuration of the alignment sensor AS will be described with reference to FIG. As shown in FIG. 2, the alignment sensor AS includes a light source 41, a collimator lens 42, a beam splitter 44,
Mirror 46, objective lens 48, condenser lens 50, index plate 52, first relay lens 54, beam splitter 56,
X-axis second relay lens 58X, X-axis imaging element 60X composed of a two-dimensional CCD, Y-axis second relay lens 58
Y, and a Y-axis imaging device 60Y composed of a two-dimensional CCD. Here, each component of the alignment sensor AS will be described together with its operation.

The light source 41 is a non-photosensitive light which does not expose the photoresist on the wafer and emits light having a broad wavelength distribution having a certain bandwidth (for example, about 200 nm). In this case, a halogen lamp is used. Used. This uses broadband illumination light to prevent a decrease in mark detection accuracy due to thin-film interference in the resist layer. This is particularly important when an image processing type alignment sensor such as the alignment sensor AS is used.

Illumination light from a light source 41 is transmitted through a collimator lens 42, a beam splitter 44, a mirror 46, and an objective lens 48 to an alignment mark MA or MB (see FIG. 4;
A "). Then, the reflected light from the alignment mark MA is irradiated on the index plate 52 via the objective lens 48, the mirror 46, the beam splitter 44, and the condenser lens 50, and the image of the alignment mark MA is formed on the index plate 52. Is done.

The light transmitted through the index plate 52 is directed to the beam splitter 56 via the first relay lens 54, and the light transmitted through the beam splitter 56 is transmitted to the X-axis imaging device 60X by the second X-axis relay lens 58X. The light focused on the imaging surface and reflected by the beam splitter 56 is focused on the imaging surface of the Y-axis imaging element 60Y by the second Y-axis relay lens 58Y. On the imaging surfaces of the imaging elements 60X and 60Y, an image of the alignment mark MA and an image of the index mark on the index plate 52 are formed so as to overlap each other. The imaging signals (DS) of the imaging devices 60X and 60Y are both supplied to the alignment control unit 16. The alignment control unit 16 is configured to include an image processing circuit and a control CPU. The alignment control unit 16 is also supplied with the measurement values of the laser interferometer 26 (see FIG. 1). The alignment control unit 16 has a function of calculating the position of the alignment mark MA on the stage coordinate system based on the image signal DS from the alignment sensor AS and the measurement value of the laser interferometer 26.

FIG. 3 shows an example of a pattern on the index plate 52 of FIG. 2. In FIG. 3, an image MAP of a cross-shaped alignment mark MA is formed at the center, and the image MAP of this image MAP is formed. Orthogonal linear pattern images MAXP and MA
The XP direction and the YP direction perpendicular to the YP are conjugate to the X direction and the Y direction of the stage coordinate system of the XY stage 20, respectively. Then, the two index marks 90A are sandwiched between the alignment mark images MAP in the XP direction.
And 90B, and two index marks 92A and 92B sandwich the alignment mark image MAP in the YP direction.
B is formed.

In this case, the index marks 90A,
90B and a detection area 94 surrounding the linear pattern image MAXP
An image in X is picked up by the X-axis image sensor 60X in FIG.
An image in the detection area 94Y surrounding the index marks 92A and 92B and the linear pattern image MAYP in the P direction is picked up by the Y-axis image sensor 60Y in FIG. Further, the scanning direction when reading the photoelectric conversion signal from each pixel of the imaging elements 60X and 60Y is set to the XP direction and the YP direction, respectively, and by processing the imaging signals of the imaging elements 60X and 60Y,
Alignment mark image MAP and index marks 90A, 90
It is possible to determine the amount of positional deviation between B and 92A, 92B in the XP and YP directions. Therefore, the alignment control unit 16 determines the stage coordinate system of the alignment mark MA from the positional relationship between the image of the alignment mark MA on the wafer W and the index mark on the index plate 52 and the measurement result of the laser interferometer 22 at that time. The coordinates on (X, Y) are obtained, and the coordinate values thus measured are supplied to the main controller 28.

Next, the alignment operation and the exposure operation in the projection exposure apparatus 100 according to the present embodiment configured as described above will be described by taking as an example the exposure of the second layer of the wafer W in the shot arrangement as shown in FIG. A description is given below.

First, a wafer W to be exposed this time is loaded onto a wafer holder (not shown) on the substrate table 18 by a wafer loader (not shown). In each shot area of the wafer W, a circuit pattern has already been formed by exposing the previous layer. Further, as shown in FIG. 4, each shot area S _n (n = 1, 2,
3,..., 32) have a first position near the center of the shot.
Cross-shaped alignment mark MA _n (n
= 1,2,3, ... 32) is formed, also the alignment mark MB _n-1 ~MB _n-4 of the cross-shaped as the second mark in the four corners of each shot area S _n (n = 1 ,
2, 3,... 32) are formed. It is also assumed that the alignment of the reticle R has been completed, and the amounts of displacement of the reticle R in the X, Y, and rotational directions with respect to the rectangular coordinates defined by a reticle interferometer (not shown) are substantially zero.

The alignment mark MA _n is designed at the center (reference point) of each shot area in design. Moreover, the design and the alignment mark MB _n-1 and MB _n-4 is formed in a position where the point symmetry with respect to the center (reference point) of each shot area, as well as MB _n-2 and MB _n-3 is In terms of design, it is formed at a point-symmetric position with respect to the center (reference point) of each shot area. Alignment mark M
The design value of the distance from the reference point of B _{n-1 to} MB _n-4 is set to a desired value (this value is known).

Next, the main controller 28 sets the origin of the wafer W (pre-alignment). After that, JP
EGA as disclosed in detail in US Pat.
(Enhanced Global Alignment)
Alignment marks MA are set as sample shots (hereinafter, referred to as “EGA shots”) for calculating a plurality of specific shot areas for shot array coordinate parameters, and correction parameters are calculated for other specific multiple shot areas. Sample shots for
The alignment mark MB which is referred to as “multipoint alignment shot in shot” is measured.

Here, as an example, in the case of the wafer W in FIG. 4, eight shot areas (S ₂ , S ₄ , S ₅ , S ₁₆ , S ₁₇ , S ₂₈ , S ₂₈₎ which are meshed as EGA shots
₂₉ , S ₃₁ ), and 16 shot areas (S ₆ , S ₇ , S ₈ , S ₉ , S ₁₂ , S ₁₃ , S ₁₄ ) which are subjected to multi-point alignment shots and are hatched. ,
_{S 15, S 18, S 19} , S 20, S 21, S 24, S 25, S 26, S
₂₇ ) The case where (2) is selected will be described.

In this case, for the eight EGA shots, the alignment marks MA are set in the same manner as in the conventional EGA.
_m (m = 2, 4, 5, 16, 17, 28, 29, 31)
Are sequentially measured. That is, the main controller 26, by driving the XY stage 20 via drive system 22 while monitoring the measurement values of laser interferometer 26 based on the coordinate values of the design of the alignment marks MA _m, the alignment sensor AS detection area (the index frame) in the first alignment marks MA, for example, to position the MA _2. Alignment control unit 16 calculates the XY coordinate position of the alignment mark MA ₂ on the stage coordinate system on the basis of the measurement values from the image signal and the laser interferometer 26 from the alignment sensor AS, the main control this calculation result Supply to device 28. Similarly, main controller 28 and alignment controller 16 control remaining alignment marks MA ₄ and MA ₄ .
₅ , XY coordinates of MA ₁₆ , MA ₁₇ , MA ₂₈ , MA ₂₉ , and MA ₃₁ are measured. There is no fundamental difference in this respect.

On the other hand, with respect to the 16 multipoint alignment shots in the shot, in this embodiment, as in the conventional case, the multipoint alignment shot S _p in each shot (p = 6,
7, 8, 9, 12, 13, 14, 15, 18, 19, 2,
Instead of measuring all the alignment marks MB in 0, 21, 24, 25, 26, 27), only the alignment marks MB in the circles A to D in FIG. 4 are measured. Hereinafter, this will be described.

FIG. 5 shows the alignment marks MB (MB _6-4 , MB _7-3 , MB _13-1 , MB _12-2 ) in the circle A in FIG.
Is shown. First, the first main controller 28, to position the alignment mark MB _6-4 above and the detection region of the alignment sensor AS in a similar manner. The relative position of the alignment sensor AS with respect to the wafer W in the state where the positioning has been performed is indicated by reference sign AS1 in FIG. In this state, main controller 28 controls alignment sensor AS and alignment controller 16
Is used to measure the position of the alignment mark MB _{6-4 on} the stage coordinate system as described above. Next, main controller 28 positions alignment mark _MB7-3 in the detection area of alignment sensor AS (reference numeral AS in FIG. 5).
2), and measure the position of the alignment mark _{MB7-3 on} the stage coordinate system. Hereinafter, in FIG.
As indicated by AS4, the alignment marks MB
_{13 -1,} to position the alignment mark MB _12-2 within the detection region of the alignment sensor AS, the alignment mark MB _13-1, to measure the position on the stage coordinate system of the alignment mark MB _12-2.

In this case, as alignment marks MB to be measured, as shown in FIG. 5, four alignment marks MB _6-4 , MB provided in the vicinity of adjacent portions where four shot areas are adjacent to each other. _7-3 , MB _13-1 ,
Since the MB _12-2 is selected, the moving amount of the XY stage 20 during the measurement is small, and although the position of the four marks is measured separately, it is slightly shorter than the measurement time for one mark. Only time is required, and the measurement time can be remarkably reduced as compared with the position measurement of four alignment marks in the target shot in the conventional multi-point intra-shot alignment.

FIG. 6 shows the alignment mark MB _6-4 , alignment mark MB _7-3 , alignment mark MB _13-1 , alignment mark MB obtained by the above measurement.
An example of the positional deviation from the design position of _12-2 is indicated by an arrow. Thereafter, similarly, the position measurement of each of the four alignment marks MB in the circles B, C, and D in FIG. 4 is performed.

Then, as described above, the eight EGAs
When the actual measurement of the alignment mark positions of the shot and the multipoint alignment shots within the 16 shots is completed, the main controller 28 uses the measurement results of the eight alignment marks MA as described in JP-A-61-44429. The shot arrangement coordinates are obtained by statistical processing using the least squares method disclosed in (1). When obtaining the shot arrangement coordinates, the measurement results of the eight alignment marks MA and the design positions of the respective alignment marks MA (the design positions are the design arrangement coordinates of the reference position of each shot area). The parameters for determining the shot arrangement coordinates are determined so that the difference between the reference arrangement position and the reference arrangement coordinates (determined based on the designed relative arrangement coordinates of each alignment mark MA) is minimized on average. However, in the case of the wafer W of FIG. 4, since the alignment marks MA are arranged at the reference positions in each shot area, the relative arrangement coordinates of the alignment marks MA with respect to the reference positions are (0, 0). In practice, there is no need to consider.

In the main controller 28, the measurement results of the 16 alignment marks MB and the design positions of the marks (the design positions correspond to the design array coordinates of the reference positions of the shot areas and the reference positions). Statistical processing is performed such that the positional deviation from the position relative to the position of the alignment mark MB is determined on the basis of the design relative arrangement coordinates), and rotation and shape error (shot error) of the reticle R in each shot area are performed. An intra-shot correction parameter for determining the orthogonality error and magnification error of the area, and a rotation and shape error of the shot area with respect to the reticle R determined by the parameter are obtained.

In calculating the above-mentioned intra-shot correction parameters, each parameter of a determinant including ten parameters is determined by the least square method, as disclosed in detail in JP-A-6-275496. Is also possible,
Japanese Patent Application Laid-Open No. Hei 6-275496 discloses a 10
Since the determinant including the six parameters includes six parameters for determining the shot array coordinates,
Except for these, a determinant including only four parameters is set, and the parameters of the determinant are set such that, for example, the positional deviation between the measurement result of the 16 alignment marks MB and the design position of each mark is minimized. It may be determined by performing statistical processing such that

Thereafter, main controller 28 applies an appropriate rotation to reticle R via reticle stage RST so as to correct the remaining rotation error θ of the circuit pattern on each shot area, and sets a stage coordinate system (X, Y). ) to correct the rotation of the circuit pattern on the shot area S _n for.

Next, the orthogonality error of the coordinate system on the wafer W cannot be corrected in a strict sense, but by rotating the reticle R appropriately, the error can be reduced. Therefore, main controller 28 can optimize the amount of rotation of reticle R or wafer W such that the sum of the absolute values of the rotation error, the rotation error, and the orthogonality error is minimized.

Next, the main controller 28, so as to correct the linear expansion (scaling error) to two orthogonal directions of the circuit pattern on each shot area S _n, via the imaging characteristics control device (not shown) To adjust the projection magnification of the projection optical system PL.

Next, the main controller 28 determines each shot area S on the wafer W based on the shot arrangement coordinates previously obtained by EGA and the previously obtained baseline amount.
The reference point of _n is sequentially aligned with a predetermined position in the exposure field of the projection optical system PL, and the pattern of the reticle R is projected and exposed on the shot area. Then, after exposing all the shot areas on the wafer W,
Processing such as development of the wafer W is performed.

In this case, the influence of the expansion and contraction and rotation of the circuit pattern transferred to each shot area is suppressed to a small value.
The circuit pattern of each shot area on the wafer W and the projected image of the pattern of the reticle R can be superimposed with higher accuracy.

In the above description, the case where two-dimensional cross-shaped marks are used as the alignment marks MA and MB has been described. However, the line-and-space pattern arranged at a predetermined pitch in the X and Y directions has been described. It goes without saying that other marks such as marks formed may be used.

As described above, according to the present embodiment,
Since the statistical processing (EGA calculation) for calculating the parameters for determining the shot arrangement coordinates and the statistical processing for calculating the in-shot correction parameters are performed separately, the EGA shot and the in-shot multipoint alignment are performed. It is possible to determine the shot and separately, as a result,
It is not necessary to detect the positions of a plurality of alignment marks in an EGA sample shot, and for multipoint alignment within a shot, it is only necessary to detect the position of each alignment mark from a plurality of different shot areas.

Therefore, even if 16 alignment marks are selected as alignment marks for multi-point alignment within a shot, for example, circle A in FIG.
By selecting the marks located in B, C and D,
Since it is possible to measure four marks in a time slightly longer than the measurement time of one mark, as a result, the mark 4
The position measurement of the 16 alignment marks can be performed in a slightly longer time than the measurement time for each. Therefore, in the present embodiment, a plurality of alignment marks (second marks) arranged in the vicinity of an adjacent part in a shot area adjacent to each other which has been selected in advance in the time required for the conventional EGA (one point measurement within a shot). By increasing the measurement time for measuring the position of the MB, it is possible to accurately determine not only the coordinates of the shot array but also the rotation error of the shot area with respect to the reticle pattern, the shape error of the shot, and the like. As a result, multi-point intra-shot alignment with high accuracy and high throughput can be performed.

In the above embodiment, when measuring the positions of the plurality of alignment marks MB arranged near the adjacent portions in the shot areas adjacent to each other, the wafer W
Has been described, and each mark MB is sequentially measured, but the present invention is not limited to this. The position measurement of a plurality of marks MB in shot areas adjacent to each other is performed by using the wafer W May be performed in the stationary state. That is, a maximum of four marks MB may be measured simultaneously. Such simultaneous measurement, for example,
This can be realized by preparing a plurality of alignment sensors AS having a thin tip shape so that they do not interfere with each other. Alternatively, this can be realized by using a low-magnification alignment microscope having a wide image field of the objective lens and having a plurality of detection criteria corresponding to each mark MB in the image field. By doing so, the measurement time is further shortened, and the measurement is performed in a state where the wafer W is stationary, so that it is possible to prevent the occurrence of measurement errors due to air fluctuations of the laser interferometer 26 and the like.

In the above embodiment, the case where completely different marks MA and MB are used as the alignment mark (first mark) for EGA measurement and the alignment mark (second mark) for multi-point alignment within a shot has been described. However, the present invention is not limited to this,
The alignment mark for GA measurement and the alignment mark for multi-point intra-shot alignment are shared, and the EGA shot and a part of the multi-point intra-shot alignment shot overlap. Therefore, measurement is performed between the EGA measurement and the intra-shot multi-point alignment measurement. Some of the marks to be made may be the same mark. In such a case, the measurement time of the entire measurement can be shortened accordingly.

In the above-described embodiment, a case has been described in which all shot areas in which alignment marks MB to be measured are present are selected as multi-point alignment shots within a shot. For example, in FIG. _6, if you select only four _{S 9, S 24, S 27} , 16 mark the same as the above embodiment in advance by the main controller 28 is predetermined sequence such that selected as the measurement object Is also good. That is, as shown in FIG. 7, for example, as shown in FIG. 7, four shots are adjacent to an arbitrary shot area S at maximum (within four circles). When the shot area S is selected, which of the above four locations the alignment mark MB should be selected as a measurement target may be arbitrarily determined in advance as a sequence.

The technical idea of selecting a plurality of marks arranged in the vicinity of the adjacent portions of a plurality of shot areas adjacent to each other as the object to be measured and utilizing the measurement result described in the above embodiment is a so-called technical idea. The present invention can also be applied to multipoint measurement within a shot in registration measurement (measurement of overlay accuracy).

Hereinafter, multipoint registration measurement within a shot in which a plurality of marks arranged in the vicinity of adjacent portions of a plurality of shot regions adjacent to each other are selected as measurement targets will be described with reference to FIG.

In FIG. 8, a dotted frame (large square)
Indicates a shot area of the first layer composed of the existing pattern formed on the substrate W in the exposure step of the first layer, and a solid line frame (large square) indicates the shot area of the substrate W in the subsequent exposure step of the second layer. 2 shows a shot area of the second layer composed of a pattern formed so as to overlap the shot area of the first layer. Also,
A small square with a solid line indicates the first measurement mark M1 exposed to each shot area S simultaneously with the pattern of the second layer, and a dotted square slightly larger than the first measurement mark M1 is formed simultaneously with the existing pattern in each shot area. 2 shows the performed second measurement mark M2. The designed relative positional relationship between the center of each corresponding first measurement mark M1 and the center of the second measurement mark M2 is a desired relationship (here, it is assumed that they match).

In performing the registration measurement, for example, a plurality of first measurement marks M1 and a second measurement mark M2 corresponding to each of the first measurement marks M1 are positioned in a circle in FIG. Relative positional relationship between four first measurement marks M1a, M1b, M1c, M1d arranged in the vicinity of an adjacent part where four shot areas are adjacent, and corresponding second measurement marks M2a, M2b, M2c, M2d Is measured by a registration measurement device (not shown).

Here, as the registration measuring device, a microscope of the same image processing system as that of the alignment sensor AS described above is used, but in this case, the index mark may not be provided. As a result of this measurement, a shift amount as shown by each arrow in the circle is obtained.

Then, by a computer (not shown), 4
The two shift amounts are statistically processed, and the overlay accuracy of the shot areas is calculated.

As is apparent from FIG. 8, the four first measurement marks M1 (M1a to M1d) to be measured and the second measurement marks M2 (M2a to M2d) corresponding to the four first measurement marks M1 (M1a to M1d).
Are arranged in the vicinity of the adjacent portions in the four mutually adjacent shot areas selected in advance, so that the relative positional relationship between the first measurement mark M1 and the second measurement mark M2 is measured. The movement of the substrate W is slight, and the measurement does not take much time. Therefore, it is possible to measure the overlay accuracy with high accuracy and high throughput.

Of course, if a registration measurement device having a large detection area is used, four sets of first measurement marks M
It is also possible to simultaneously measure the relative positional relationship between the first and second measurement marks M2. By doing so, it is possible to further improve the throughput.

[0077]

As described above, according to the first aspect of the invention, there is an excellent effect that high-precision and high-throughput multipoint alignment within a shot can be performed.

According to the second aspect of the present invention, it is possible to further improve the throughput in addition to the effect of the first aspect of the present invention.

According to the third aspect of the present invention, there is provided a method of measuring overlay accuracy with high accuracy and high throughput.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a schematic configuration of a projection exposure apparatus according to an embodiment.

FIG. 2 is a diagram showing a specific configuration example of the alignment sensor of FIG. 1;

FIG. 3 is an enlarged view showing an arrangement of an alignment mark image and an index mark on the index plate of FIG. 2;

FIG. 4 is a view for explaining an alignment operation in the projection exposure apparatus according to one embodiment, and is a view showing an example of a shot arrangement on a wafer.

5 is a diagram showing a state of measurement of an alignment mark MB in a circle A in FIG.

FIG. 6 shows alignment marks MB _6-4 obtained by measurement,
FIG. 14 is a diagram illustrating an example of a positional deviation of MB _7-3 , MB _13-1 , and MB _12-2 with respect to a design position.

FIG. 7 is a diagram for explaining a method of selecting an alignment mark MB to be measured for a shot area selected as a multi-point alignment shot within a shot.

FIG. 8 is a diagram for describing multipoint registration measurement within a shot in which a plurality of marks arranged near an adjacent portion of a plurality of shot regions adjacent to each other are selected as measurement targets.

FIG. 9 is a diagram for explaining a conventional example.

[Explanation of symbols]

 W wafer (substrate) R reticle (mask) S shot area MA alignment mark (first mark) MB alignment mark (second mark) M1 first measurement mark M2 second measurement mark

Claims (3)

[Claims] A plurality of shot areas arranged on a substrate and onto which a pattern on a mask is transferred are each aligned with a predetermined transfer position in a stationary coordinate system defining a movement position of the substrate. An alignment method for determining array coordinates of the plurality of shot areas on the static coordinate system and at least one of a rotation error and a shape error of each shot area with respect to the pattern on the mask, Of a plurality of position detection marks, each of which is arranged in a predetermined relative positional relationship with respect to a reference position within the plurality of shot areas, at least one of the first marks is arranged in a plurality of shot areas selected in advance. A first process for measuring a position and a position of a plurality of second marks arranged in the vicinity of an adjacent portion in a shot region adjacent to each other which is selected in advance; Using the measurement result of the first step, the designed array coordinates of the reference position in each of the shot areas, and the designed relative array coordinates of the first mark and the second mark with respect to the reference position, respectively; A second step of performing predetermined statistical processing to obtain array coordinates of the plurality of shot areas on the stationary coordinate system and at least one of a rotation error and a shape error of each shot area with respect to the pattern on the mask; An alignment method including: 2. The alignment method according to claim 1, wherein the position measurement of the plurality of second marks in the mutually adjacent shot areas in the first step is performed while the substrate is stationary. 3. A first measurement mark which is simultaneously exposed to a new pattern in a plurality of shot areas on a substrate, and a first measurement mark is formed in each of said shot areas at the same time as an existing pattern,
An overlay accuracy measurement method for measuring a relative positional relationship with a second measurement mark whose design relative positional relationship with the first measurement mark is known, wherein the plurality of first measurement marks and each first measurement are measured. From among the second measurement marks corresponding to the marks, the relative positional relationship between the plurality of first measurement marks arranged near the adjacent portions in the mutually adjacent shot areas and the second measurement marks corresponding thereto is measured. A superposition accuracy measuring method including a first step and a second step of performing predetermined statistical processing using the measurement result of the first step.