JP7229637B2 - Exposure device and exposure method - Google Patents

Description

The present invention relates to a projection exposure apparatus for forming a pattern formed on a reticle or the like on a substrate, and more particularly to alignment (positioning) when patterns are superimposed and transferred onto the substrate.

Most of devices such as semiconductor elements, liquid crystal display elements, package substrates, etc. manufactured using a projection exposure apparatus have a multi-layer structure, and a pattern is superimposed and transferred onto a substrate. When arranging the same pattern on a substrate at a predetermined pitch, a plurality of identical patterns are arranged on a photomask and the patterns are simultaneously exposed in one shot in order to improve throughput.

When patterns are superimposed and transferred, it is necessary to perform alignment so as to be within the allowable registration error range. Therefore, every time the exposure process is repeated, a one-shot exposure area is selected to suppress the overlay error.

For example, at each exposure step, from among a reticle in which a plurality of fields with different numbers of pattern arrangements are formed, a field that becomes the maximum exposure area per shot within a range that satisfies a predetermined alignment accuracy is selected, and patterning is performed ( See Patent Document 1). Also, a field corresponding to the deformation of the substrate is selected from a reticle in which a plurality of fields with different shapes of exposure regions are formed, and patterning is performed (see Patent Document 2).

Also, every time the pattern is superimposed, the alignment of the substrate as a whole is corrected so that the position of the alignment mark on the substrate is gradually moved to the designed position, thereby aligning the shot area in the uppermost layer with the designed position. (See Patent Document 3).

JP 2017-44721 A JP 2016-51712 A International Publication No. 2012/124517

In the case of a substrate on which a large number of chips, packages, etc. are two-dimensionally arranged (for example, a FO-WLP (fan-out-wafer level package) substrate in which silicon chips are mounted on a carrier), the number of chips is enormous. In this case, since random chip arrangement errors occur due to chip mounting accuracy, it is necessary to perform alignment by detecting the positions of alignment marks provided for each chip. This takes a lot of time and reduces throughput. Therefore, it is necessary to measure the positions of the alignment marks in advance using a measuring device independent of the exposure device.

On the other hand, expansion and contraction and deformation of the substrate occur in each process, and the degree of expansion and contraction and deformation also differs. Therefore, the position of the alignment mark measured prior to the exposure apparatus may be different when the alignment measurement is performed later by the exposure apparatus, and there is a possibility that the selection of the one-shot exposure area and the alignment correction will be erroneously performed. be.

Therefore, it is required to overlap and transfer patterns while suppressing the amount of alignment error by selecting an appropriate exposure area.

An exposure apparatus according to the present invention is an exposure apparatus capable of exposing a plurality of units on a substrate and capable of communicating with a measuring device for measuring the position of an alignment mark formed on each unit. A measurement unit that measures the position of the GA (global alignment) alignment mark, the position information of each unit obtained based on the alignment mark information sent from the measurement device, and the position information of the GA alignment mark obtained by the measurement unit. a residual prediction unit that predicts an alignment residual between the exposure pattern formed on the substrate and each unit based on the positional information. Then, the exposure apparatus switches the reticle field for a reticle provided with a plurality of reticle fields or exchanges the reticle based on the alignment error. A unit is, for example, a chip, and serves as an exposure target area.

As a configuration for switching reticle fields, for example, the reticle has a plurality of reticle fields that define one-shot areas of different sizes, and the exposure apparatus selects the reticle field based on the alignment residual. Adjust the number of units exposed in a shot. Alternatively, the reticle is a reticle having a plurality of reticle fields written with mutually different scaling amounts, and the exposure apparatus selects the reticle field and performs scaling correction based on the alignment residual.

The residual predictor can further predict an alignment residual using predetermined design position information for each unit. In this case, the exposure apparatus obtains a target coordinate position that is offset by a predetermined amount in a direction approaching the design position of each unit with respect to the position where the alignment error of each unit is minimized, and determines the target coordinate position on the substrate and the reticle. and are relatively aligned.

Such alignment is effective when the multi-chip package substrate area having a plurality of units is arranged in multiple planes. For example, the residual prediction unit predicts the alignment residual of each unit for each multi-chip package substrate area, and sets target coordinates so that the alignment residual is within a threshold. If the alignment error of each unit is within the threshold value, the exposure apparatus should collectively expose the multi-chip package substrate area.

On the other hand, if the alignment error of each unit is greater than the threshold, a different pattern may be exposed for each unit by replacing the reticle. Alternatively, different patterns may be exposed for each unit by switching the reticle field.

According to another aspect of the present invention, there is provided an exposure method in which a measuring device independent of an exposure device measures the position of each unit arranged on a substrate from alignment marks formed for each unit. A measurement section provided in the exposure apparatus measures the position of an alignment mark for GA (global alignment) provided on the substrate, and the position information of each unit obtained by the measurement apparatus and the GA obtained by the measurement section. Based on the positional information of the alignment marks for the substrate, an alignment error between the exposure pattern formed on the substrate and each unit is predicted, and based on the alignment error, a reticle for a reticle provided with a plurality of reticle fields. Field switching or reticle exchange.

According to the present invention, in a projection exposure apparatus, patterns can be overlapped and transferred while suppressing the amount of alignment error.

1 is a schematic block diagram of a projection exposure apparatus according to this embodiment; FIG. 3 is a diagram showing a shot arrangement formed on a substrate W; FIG. FIG. 4 is a diagram showing shot arrangement errors by dividing them into components; 4 is a diagram showing a flowchart of chip-specific arrangement error calculation processing executed by the measuring device 15. FIG. 2 is a diagram showing exposure processing including alignment performed in the projection exposure apparatus 10; FIG. FIG. 11 is a diagram showing a flowchart of exposure processing including alignment in the second embodiment; 4 is a plan view showing a reticle R; FIG. FIG. 7 is a diagram showing a subroutine of step S305 of FIG. 6; FIG. 10 is a diagram showing a procedure for selecting reticle fields F1 to F4 used during exposure; It is the figure which showed the deformation|transformation condition of a one-shot exposure area|region. FIG. 4 is a diagram showing a reticle on which a plurality of mask patterns are formed; FIG. 10 is a diagram showing a multi-chip package;

Embodiments of the present invention are described below with reference to the drawings.

FIG. 1 is a schematic block diagram of the projection exposure apparatus of this embodiment. Here, an exposure process is performed in which patterns are superimposed on a plurality of layers.

The projection exposure apparatus 10 is an exposure apparatus that transfers a mask pattern formed on a reticle R as a photomask onto a substrate (work substrate) W according to a step-and-repeat method. It has The reticle R is made of a quartz material or the like, and has a mask pattern with a light shielding area. As the substrate W, a substrate made of silicon, ceramics, glass, or resin (for example, an interposer substrate) is applied here.

The illumination light emitted from the light source 20 enters the integrator 24 via the mirror 22, and the illumination light amount becomes uniform. The uniform illumination light enters the collimator lens 28 via the mirror 26 . As a result, parallel light is incident on the reticle R. As shown in FIG. The light source 20 is driven and controlled by a lamp driving section 21 .

The reticle R is mounted on the reticle stage 30 so that the mask pattern becomes the focal position of the projection optical system 34 on the light source side. A stage 30 on which the reticle R is mounted and a stage 40 on which the substrate W is mounted are defined with XYZ three-axis coordinate systems orthogonal to each other. Stage 30 is movable in the XY directions to move reticle R along the focal plane and is driven by stage drive 32 . Stage 30 is also rotatable in the XY coordinate plane. The position coordinates of stage 30 are measured here by a laser interferometer or a linear encoder (not shown).

The light transmitted through the area of the reticle R where the mask pattern is formed is projected as pattern light onto the substrate W by the projection optical system 34 . The substrate W is mounted on the substrate stage 40 so that its exposure surface coincides with the image-side focal position of the projection optical system 34 .

The stage 40 is movable in the XY directions to move the substrate W along the focal plane and is driven by a stage drive 42 . Further, the stage 40 can move in the Z-axis direction (optical axis direction of the projection optical system 34) perpendicular to the focal plane (XY direction), and can also rotate in the XY coordinate plane. The position coordinates of the stage 40 are measured by a laser interferometer (not shown) or a linear encoder.

The control unit 50 controls the stage driving units 32 and 42 to position the reticle R and the substrate W, and controls the lamp driving unit 21 . Then, an exposure operation based on the step & repeat method is executed. A memory (not shown) provided in the control unit 50 stores the mask pattern position coordinates of the reticle R, the designed position coordinates of the shot area formed on the substrate W, the step movement amount, and the like.

The alignment mark imaging unit 36 is a camera or a microscope that images an alignment mark formed on the substrate W, and drives the stage driving unit 42 to move the stage 40 to move the alignment mark provided on the substrate W before exposure. Take an image of the mark. The image processing section 38 detects the position coordinates of the alignment marks based on the image signal sent from the alignment mark imaging section 36 . Here, the positions of GA (global alignment) alignment marks provided at the four corners of the substrate W are detected.

The control unit 50 sequentially transfers the mask pattern of the reticle R to each shot area formed on the substrate W according to the step-and-repeat method. That is, the control unit 50 intermittently moves the stage 40 according to the shot area interval, and when the shot area to be exposed is positioned at the projection position of the mask pattern, the control unit 50 drives the light source 20 to project the pattern light onto the shot area. project it.

The measuring device 15 is a device for measuring alignment marks provided on the substrate W, and is connected to the projection exposure apparatus 10 so as to be able to communicate with each other. The substrate W here is composed of an FO-WLP substrate, and a large number (for example, 10,000) of silicon chips are arranged in a matrix on a carrier substrate. The substrate is not limited to the FO-WLP substrate, and may be a normal wafer substrate, printed wiring board, or the like. The chips may not be of the same type, and multiple types of chips may be arranged. Alternatively, the substrate may be a substrate on which a large number of circuit patterns (referred to as units) having a smaller area than the substrate are arranged in a matrix, and the exposure may be performed by aligning with the units as a reference. There may be one or more chips inside the unit. The measuring device 15 measures the positions of alignment marks provided on each chip before patterning by the exposure device 10 .

The measurement device 15 includes a substrate stage on which the substrate W is placed, a camera that captures an image of the alignment mark, a scanning mechanism that moves the substrate relative to the camera, a measurement unit that measures the amount of relative movement of the substrate, and a camera. An image processing unit that measures the position of the alignment mark from an image obtained by imaging, a data management unit that records the position information of the measured alignment mark as a file, and a controller that controls them (here, any not shown).

The measuring device 15 transmits to the data server 60 a file regarding the positional information of the alignment marks of each chip. The data server 60 records and manages the file in association with the ID of the substrate W to be measured.

The projection exposure apparatus 10 performs alignment when transferring (exposing) a mask pattern to each layer. That is, an alignment error, which is an alignment error of the shot areas, is detected, and alignment between the shot areas of the substrate W and the projection area of the mask pattern is performed in advance.

In this embodiment, the projection exposure apparatus 10 calculates the amount of positional deviation peculiar to each chip from the position of the alignment mark of each chip measured by the measuring device 15. Using the measured positions of the GA alignment marks, the exposure position of each chip during the exposure process is determined. The projection exposure apparatus 10 performs alignment exposure according to the die-by-die (hereinafter referred to as D/D) method according to the determined exposure position of each chip. This will be described in detail below.

FIG. 2 is a diagram showing a shot arrangement formed on the substrate W. As shown in FIG. FIG. 3 is a diagram showing the distortion of the shot arrangement caused by the deformation of the substrate W or the like.

As shown in FIG. 2, the substrate W is formed with a lower layer pattern in which chips CP are arranged in a matrix at regular intervals in accordance with a grid defined by an XY coordinate system. Here, each chip CP corresponds to a shot area, and the mask pattern formed on the reticle R is formed over the chip CP (hereinafter also referred to as shot area). Alignment marks AM for alignment are formed at arbitrary positions (central positions of left and right ends in FIG. 2) along the arrangement of the shot areas CP. The alignment mark AM for alignment may be provided at a position near the shot area CP (for example, on the scribe line). Further, some of the alignment marks AM for alignment may also be used as alignment marks for GA.

FIG. 3 shows a state in which the alignment of the shot areas CP is disrupted due to the deformation of the substrate W. As shown in FIG. When the substrate W is a wafer, the amount of deformation is not so large, but it is not negligible when compared with the allowable error range for overlaying the underlying pattern. Moreover, when the board W is a printed board or an interposer board, the deformation of the board W is further increased. Therefore, the position of each chip CP deviates from the designed position.

When the substrate W is an FO-WLP substrate, random chip alignment errors (called die shift) occur due to chip mounting accuracy. On the other hand, the substrate W undergoes expansion, contraction and deformation (collectively referred to as substrate deformation) each time it undergoes a process. Therefore, the position of the alignment mark AM of each chip CP during exposure differs from the position of the alignment mark AM measured by the measuring device 15 .

In particular, when patterning is performed on a plurality of layers, each time a process including exposure (exposure, etching, heat treatment, etc.) is repeated, the degree of expansion and contraction and deformation of the substrate W changes, and the alignment mark measured by the measurement device 15 changes. Positional deviation from AM increases.

FIG. 3 is a diagram showing the components of the shot arrangement error. Chip arrangement error components due to die shift (defined as die shift amounts) have no regularity in error amounts between chip CPs. This non-regular arrangement error is defined as a specific positional deviation amount. The inherent positional deviation amount can be regarded as a non-linear (non-linear) error component regardless of the deformation and expansion/contraction of the substrate W as a whole. On the other hand, the chip arrangement error component due to the deformation, expansion and contraction of the substrate W is an arrangement error that occurs with a common law among the chips CP according to the degree of deformation of the entire substrate W, so it is regarded as a linear error component. be able to.

Therefore, first, from the chip array error amount between the position of the alignment mark AM of each chip CP measured by the measuring device 15 and the designed position, the linear array error amount is removed, and the unique position of each chip is calculated. Extract the amount of deviation. Then, in the exposure apparatus 10, the linear chip arrangement error amount obtained from the GA alignment mark is obtained, and the exposure position of each chip in the exposure process is estimated by combining the inherent positional deviation amount of each chip.

FIG. 4 is a diagram showing a flowchart of the chip-specific arrangement error calculation process executed by the measuring device 15. As shown in FIG.

When the positions of the alignment marks provided on each chip of the substrate W are measured (S101), the difference (X, Y, θ) between the position information of the alignment marks on each chip and the design position information is
is calculated (S102). Then, from the obtained chip arrangement error amount, a linear component is extracted using the method of least squares or the like (S103), and by removing this, the X, Y, and .theta. (S104).

FIG. 5 is a diagram showing exposure processing including alignment performed in the projection exposure apparatus 10. As shown in FIG.

The positions of the GA alignment marks provided on the substrate W (for example, the positions of the alignment marks provided at the four corners of the substrate) are measured, and the difference from the design position is divided into the X component, the Y component, and the θ component. Calculate (S201). On the other hand, the data of the unique positional deviation amount of each chip obtained by the measuring device 15 is read from the data server 60 (S202).

The positional deviation amount of the GA alignment mark measured by the exposure apparatus 10 is a linear positional deviation amount (that is, the X, Y, and θ deviation amounts when the substrate is placed on the exposure apparatus, the expansion and contraction of the substrate, and the like. Since it can be regarded as the sum of deformation amounts such as changes in orthogonality, the deformation amount of the entire substrate W in the exposure process can be calculated from this. Then, by the GA alignment technique, it is possible to calculate the amount of positional deviation of the X, Y, and θ components at the exposure position of each chip obtained from the amount of deformation of the entire substrate W (calculated from the amount of deformation of the entire substrate W). The amount of positional deviation at the exposure position of each chip is also defined as the amount of linear positional deviation). However, it is assumed that the deformation and expansion/contraction of the substrate W occur substantially uniformly over the entire substrate or with respect to the exposure target area.

Then, the exposure position of each chip is calculated based on the linear positional deviation amount and the intrinsic positional deviation amount of each chip (S203). That is, the exposure position (exposure map) of each chip during the exposure process is calculated by adding the linear positional deviation amount and the inherent positional deviation amount to the designed exposure position. Although the projection exposure apparatus 10 does not measure the alignment marks of each chip, the exposure position of each chip is estimated by this calculation, and exposure is performed by the die-by-die method (S204).

As described above, according to this embodiment, the positions of the alignment marks of the chips arranged in a matrix on the substrate W are measured in advance by the measurement device 15 independent of the projection exposure apparatus 10, and the inherent positional deviation amount of each chip is measured. Ask for Then, the projection exposure apparatus 10 measures the GA alignment marks and calculates the positional deviation amount of each chip having linearity. Based on the inherent positional deviation amount and the linear positional deviation amount, the exposure position of each chip is calculated and the exposure is performed while the substrate W is mounted on the exposure apparatus 10 .

In this way, by calculating the positional deviation amount with linearity that changes with the process during the exposure process, and by calculating the inherent positional deviation amount that occurs when the chip is first mounted and does not change even after the process, before using the exposure apparatus. , a proper alignment can be performed. In particular, even when repeating patterns are superimposed on the substrate W, appropriate alignment can be performed. This can also be applied to a substrate on which elements other than chips are two-dimensionally arranged.

Instead of the measuring device 15 calculating the chip-specific positional deviation amount and outputting it to the projection exposure apparatus 10, the data server 60 or the projection exposure apparatus 10 may calculate the chip-specific positional deviation amount. Further, the GA alignment mark is not limited to one formed at a specific location on the substrate W, and a predetermined alignment mark may be selected to obtain a linear positional deviation amount. It is also possible to apply to a maskless exposure apparatus instead of a projection exposure apparatus. In this case, the exposure data may be corrected based on the obtained position of the unique positional deviation amount of the alignment mark of each chip.

Further, when exposing a predetermined layer of the substrate W, the measuring device 15 calculates the chip-specific positional deviation amount and outputs it to the projection exposure device 10, and when exposing a layer above that layer, Alignment may be performed using the chip-specific positional deviation amount that has already been calculated.

Next, a second embodiment will be described with reference to FIGS. 6 to 12. FIG. In the second embodiment, it is confirmed by simulating in advance how much the amount of alignment error occurs when patterning is performed.

FIG. 6 is a diagram showing a flowchart of exposure processing including alignment in the second embodiment. However, it shows alignment adjustment and exposure processing during a predetermined exposure process. Further, exposure is performed according to a predetermined exposure method.

As in the first embodiment, the measurement device 15 measures the position of the alignment mark provided on each chip of the substrate W, obtains the inherent positional deviation amount of each chip, and stores it in the data server 60 (S101 to S104). ). The substrate W is transported to the stage 40 of the exposure apparatus 10, the inherent positional deviation amount of each chip is read from the data server 60, and the positions of the GA alignment marks are measured (S301, S302). Then, the positional deviation amount (linear positional deviation amount) of the X, Y, and .theta. components at each chip exposure position is obtained (S303).

Then, the exposure position of each chip during the exposure process is calculated based on the linear positional deviation amount of each chip and the inherent positional deviation amount of each chip measured by the measuring device 15 (S304).

A simulation is performed to find out how much positional deviation remains when the obtained exposure position of each chip in the exposure process is exposed according to a predetermined exposure method (S305). Here, as an exposure method, an exposure method is adopted in which an appropriate reticle field is selected from among a plurality of reticle fields having different numbers of pattern arrangements for patterning. This exposure method will be described with reference to FIGS.

FIG. 7 is a plan view showing the reticle R. FIG. The reticle R is provided with four reticle fields F1 to F4 in the reticle body RB, which serves as a light shielding portion, and the same mask patterns MP are formed so that the numbers of patterns arranged are different. The numbers of pattern arrays of the reticle fields F1, F2, F3, and F4 are 1 (=2 ⁰ ), 4 (=2 ² ), 8 (=2 ³ ), and 16 (=2 ⁴ ), respectively.

When the substrate W is free from deformation and die shift, there is no arrangement error in the shot area. Therefore, the number of shots for the entire substrate W can be minimized by pattern transfer using the reticle field F4 having the largest number of pattern arrangements. However, if the substrate W is greatly deformed or the die shift is large, exposure using the reticle field F4 with the largest number of pattern arrays may cause mask patterns to be overlaid beyond the allowable overlay accuracy.

Therefore, the entire exposure target area of the substrate W is divided into a plurality of areas (hereinafter referred to as calculation areas), and the calculation areas are defined as one-shot exposure areas to perform simulation. Then, among the plurality of reticle fields F1 to F4, the reticle field in which the estimated (calculated) misregistration residual amount for each chip (pattern formation area) is equal to or less than the threshold value and the number of patterns becomes maximum. to select.

FIG. 8 is a diagram showing the subroutine of step S305 in FIG.

First, chips arranged in a predetermined one-shot exposure area determined according to the exposure method are extracted (S401). Then, the design chip center position with respect to the shot center of the one-shot exposure area is calculated, and the alignment correction value of the one-shot exposure area is calculated based on the position information of the alignment marks of each extracted chip (pattern formation area). It is calculated as X, Y, and θ components.

This alignment correction value is calculated for each one-shot exposure area (S402). Then, the amount of misalignment between the center position of each chip and the designed center position (hereinafter referred to as alignment residual amount) when alignment exposure is performed using the alignment correction value is calculated (S403).

In step S404, it is determined whether or not the alignment residual amount is equal to or less than a threshold. Here, the threshold represents the maximum allowable amount within a range in which misalignment of wiring does not affect when patterning is repeated. If the alignment residual amount is equal to or less than the threshold, it is determined to perform exposure in the previously determined one-shot area (S405). On the other hand, if the alignment residual amount is not equal to or less than the threshold, the next reticle field, that is, the one-shot exposure area is selected (S406), and steps S401 to S403 are executed to obtain the alignment residual amount.

FIG. 9 is a diagram showing a procedure for selecting reticle fields F1 to F4 used during exposure. First, a one-shot exposure area OA1 is determined in accordance with the number of patterns arranged in the reticle field F4. The alignment residual amount of the alignment mark is calculated for each of the four one-shot exposure areas OA1.

For areas where the amount of registration residual is less than or equal to the threshold, reticle field F4 is selected. On the other hand, if the alignment residual amount exceeds the threshold, the reticle field F3 is newly selected as the one-shot exposure area, and the alignment residual amount is calculated by simulation. Reticle field F2 is selected for areas where the amount of registration residual exceeds the threshold. In addition, reticle field F1 is selected for areas where the amount of registration residual exceeds the threshold.

In this way, a simulation is performed in which the reticle fields are exposed in descending order of size and the reticle fields are used to determine the amount of alignment residual error. Then, after confirming that the alignment residual amount is equal to or less than the threshold value, the one-shot exposure area is set.

Once the reticle field of the one-shot exposure area is selected for the entire substrate, alignment exposure is performed based thereon (step S306 in FIG. 6). By repeating steps S301 to S306 for each layer, patterns are superimposed and transferred onto the substrate W. FIG.

Here, an exposure method is employed in which a plurality of reticle fields with different numbers of pattern forming regions included in the one-shot exposure region are prepared. Exposure is also possible. This exposure method will be described in detail below.

FIG. 10 is a diagram showing how the one-shot exposure area is deformed. The one-shot exposure area VP1 maintains a rectangular shape without substrate deformation, and has a reference one-shot exposure area shape. When linear expansion/contraction occurs in the substrate W, the reference one-shot exposure area VP1 maintains its rectangular shape, and only its dimensions change. One-shot exposure areas VP2 and VP3 show the shape of linear deformation.

On the other hand, if the sides facing each other in the +/− direction with respect to the X direction/Y direction have a deviation of the same degree of orthogonality, the one-shot exposure area changes into a parallelogram shape. The one-shot exposure region VP4 has a parallelogram (rhombus) shape with a deviation in orthogonality in the -X direction, and the one-shot exposure region VP6 has a deviation in orthogonality in the -Y direction. there is Also, the one-shot exposure areas VP5 and VP7 exhibit a non-rectangular shape with a larger deviation in orthogonality.

Further, the reference one-shot exposure area VP1 may be deformed into a trapezoidal one-shot exposure area VP' due to deviation in the degree of orthogonality such that the opposite sides approach each other. As described above, the deformation of the shot area from a rectangular shape to a non-rectangular shape due to the distortion of the substrate W varies.

FIG. 11 is a diagram showing a reticle on which a plurality of mask patterns are formed. Here, a plurality of mask patterns (deformed mask patterns) are formed on the reticle R in accordance with one-shot exposure areas deformed into various non-rectangular shapes. As shown in FIG. 3, eight mask patterns are formed here on the reticle R, and the field shape of each mask pattern corresponds to the deformed shape of the one-shot exposure area (however, the deformed shape is very small). Therefore, the deformation is not shown in FIG. 10 so that the deformation can be seen.).

For such an exposure method, a simulation is performed while sequentially selecting reticle fields, and it is determined whether or not the amount of alignment residual error is below a threshold value. Exposure can be performed. Note that exposure may be performed by selecting a reticle from among a plurality of reticles.

Furthermore, it is also possible to apply an array control alignment exposure method as an exposure method. In the array control alignment exposure method, the target coordinate position is calculated by moving the calculated position of the alignment mark in each pattern formation area to the designed position by a predetermined distance (for example, halfway point), and alignment based on this is calculated. perform, that is, modify the alignment correction value.

In this case, a simulation of exposure is performed based on the target coordinate position, and the alignment residual amount of each pattern formation region (chip) is calculated. In the array control alignment exposure method, the arrangement of each chip generally approaches the design value each time the exposure process is repeated for each layer. As a result of the simulation, if the alignment residual amount is not equal to or less than the threshold value, array control alignment is performed for each chip to determine the exposure position, and die-by-die exposure is performed. On the other hand, if the alignment residual amount is less than the threshold, the chip position is assumed to be the designed position, and the one-shot exposure area is expanded to expose multiple chips at once (full-shot exposure). )do.

As described above, according to the present embodiment, based on the positional information of the chip obtained by the measuring device 15, that is, the positional information of the alignment marks in each pattern forming region, and the positional information of the GA alignment marks obtained by the exposure device 10, , the exposure position of each pattern forming region is calculated during the exposure process, that is, when the substrate W is mounted on the stage.

Then, an alignment exposure performed in a reticle field or one-shot exposure area determined according to a predetermined exposure method is simulated, and an alignment residual amount for each (remaining) chip at that time is obtained for each one-shot exposure area. If the amount of registration residual is less than or equal to the threshold, then decide to expose in that one-shot exposure area, otherwise choose the next reticle field to simulate or do a die-by-die exposure. .

By performing a simulation in advance based on the alignment mark position information obtained by the measuring device 15 and the exposure device 10 at different timings, it is possible to check whether the amount of alignment error is within a desired range before exposure. It is possible to suppress the occurrence of registration errors beyond the expected range.

The simulation may be performed according to an exposure method other than the exposure method described above. Alternatively, in the array control alignment exposure method, instead of die-by-die exposure, exposure by switching the reticle field described above may be performed. As for the substrate, the present invention can be applied not only to substrates having chip arrays, but also to substrates such as package substrates in which various pattern formation regions are two-dimensionally arranged.

As shown in FIG. 12, the substrate may be a substrate on which multiple multi-chip package regions, each of which constitutes one package F with a plurality of units (F1 to F3), are arranged. In this case, the alignment residual of each unit is predicted by simulation for each multi-chip package substrate area, and the target coordinates are set so that the alignment residual is within the threshold. When the alignment residual error of each unit is within the residual threshold value, the time required for alignment and exposure can be shortened by collectively exposing the multi-chip package area.

In a multi-chip package, different types of chips may be arranged in the package, and the units may also expose different patterns in accordance with the chips. It is possible to prepare a different pattern for each reticle field of the reticle and expose patterns of different units by switching the reticle field. Alternatively, it is possible to provide different patterns for each reticle and expose patterns of different units by exchanging the reticle itself.

REFERENCE SIGNS LIST 10 projection exposure apparatus 15 measuring apparatus 40 stage 42 stage driving section 50 control section W substrate AM alignment mark R reticle

Claims (11)

An exposure apparatus capable of exposing a plurality of units on a substrate and capable of communicating with a measuring apparatus for measuring positions of alignment marks formed on each unit,
a measurement unit that measures the position of a GA (global alignment) alignment mark provided on the substrate;
position information of each unit obtained from the position of the alignment mark for each unit measured by the measuring device before position information measurement of the alignment mark for the GA by the measuring unit; and the position information for the GA obtained by the measuring unit Based on the position information of the alignment mark, the exposure position of each unit is calculated during the exposure process, and the alignment residual error between the exposure pattern formed on the substrate and each unit when the exposure is performed according to the determined exposure method. and a residual prediction unit that predicts
An exposure apparatus that switches reticle fields for a reticle provided with a plurality of reticle fields or exchanges the reticle based on an alignment error. The reticle is a reticle having a plurality of reticle fields defining one-shot areas of different sizes,
2. An exposure apparatus according to claim 1, wherein said exposure apparatus selects said reticle field based on an alignment error and adjusts the number of units to be exposed in one shot. The reticle is a reticle having a plurality of reticle fields drawn with mutually different scaling amounts,
2. An exposure apparatus according to claim 1, wherein said exposure apparatus selects said reticle field based on said registration error and performs scaling correction. 2. An exposure apparatus according to claim 1, wherein said residual prediction section further predicts an alignment residual using predetermined design position information of each unit. A target coordinate position offset by a predetermined amount in a direction approaching the design position of each unit is obtained with respect to the position where the alignment error of each unit is minimum, and the target coordinate position on the substrate and the reticle are relatively aligned. 5. The exposure apparatus according to claim 4, wherein: 6. An exposure apparatus according to claim 4, wherein said substrate is a substrate on which multiple multi-chip package substrate areas having a plurality of units are arranged. The residual prediction unit predicts an alignment residual of each unit for each multi-chip package substrate area, and sets target coordinates so that the alignment residual is within a threshold value. The exposure apparatus according to claim 6. 8. The exposure apparatus according to claim 7, wherein the multi-chip package substrate area is collectively exposed when the alignment error of each unit is within a threshold value. 8. An exposure apparatus according to claim 7, wherein when the alignment error of each unit becomes larger than a threshold, different patterns are exposed for each unit by exchanging the reticle. 8. An exposure apparatus according to claim 7, wherein when the alignment error of each unit is greater than a threshold, different patterns are exposed for each unit by switching the reticle field. The position of each unit arranged on the substrate is measured from the alignment mark formed for each unit by a measuring device that measures the position of the alignment mark and is independent of the exposure device and can communicate with the exposure device. death,
After measuring the position of each unit by the measuring device, the position of a GA (global alignment) alignment mark provided on the substrate is measured by a measuring unit provided in the exposure device,
Based on the positional information of each unit obtained by the measuring device and the positional information of the GA alignment mark obtained by the measuring unit, the exposure position of each unit during the exposure process is calculated, and according to the determined exposure method. Predicting an alignment residual between each unit and an exposure pattern formed on the substrate when exposed ,
An exposure method comprising switching a reticle field for a reticle provided with a plurality of reticle fields or exchanging the reticle based on an alignment error.

Images