JP2007324199A - Exposure system and exposure method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an exposure system and an exposure method with which a wafer where overlay deviation suddenly occurs can securely be detected without dropping throughput. <P>SOLUTION: An arrangement error measuring part 21 measures an arrangement error belonging to a specified layer that is already formed on a substrate. An element component calculator 22 decomposes the measured arrangement error into a plurality of element components. A fluctuation width calculator 23 calculates at least fluctuation width of one element component obtained for a plurality of substrates, preferably, the substrates between an orthogonal degree component and a scaling component. A comparator 3 compares calculated fluctuation width of the element component, and fluctuation permission width which is previously set for the element component and outputs comparison results. A decision part 4 decides presence of exposure processing abnormality based on the comparison results output from the comparator 3. Thus, a lot where sudden overlay deviation occurs can be detected in the exposure process. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an exposure system and an exposure method for transferring a pattern to a workpiece in a manufacturing process of a semiconductor device or the like, and more particularly, to an exposure system and an exposure method for performing an exposure process on a lot basis by a step-and-repeat method.

  In an exposure process in a manufacturing process of a semiconductor device such as a semiconductor integrated circuit, with the miniaturization of an element pattern, it is required to increase the accuracy of alignment in which an upper element pattern is superimposed on a lower element pattern. The alignment is performed by measuring the arrangement state of pattern layer shots (exposure units) already formed on the wafer and aligning the shot position of the pattern layer to be exposed with the arrangement state of the pattern layers on the wafer. Is done. In order to enable such alignment, a wafer alignment mark (hereinafter simply referred to as an alignment mark) is formed on each shot belonging to the pattern layer on the wafer.

  As the alignment method, die-by-die alignment and global alignment are well known. Die-by-die alignment is a technique for aligning shots to be exposed for each shot belonging to a specific pattern layer already formed on the wafer. That is, an exposure apparatus that employs die-by-die alignment first measures the coordinates of alignment marks formed on the shot of the specific pattern layer corresponding to the shot to be exposed. Based on the measurement result, the exposure apparatus determines the position on the wafer of the shot to be exposed. On the other hand, global alignment is a technique for performing alignment based on the arrangement state of a plurality of specific shots selected from all shots belonging to a specific pattern layer already formed on the wafer. That is, an exposure apparatus that employs global alignment first measures the coordinates of alignment marks formed on the specific shot. Based on the measurement result, the exposure apparatus predicts the arrangement state of all shots on the wafer. Based on the predicted result, the exposure apparatus determines the position on the wafer of the shot to be exposed. At this time, the number of specific shots selected on the wafer is different depending on the type of process and the exposure accuracy depending on the type of pattern layer to be exposed and the required alignment accuracy. It is often about 12 points.

  Global alignment can be performed in a shorter time than die-by-die alignment because the number of alignment mark measurement points is small. In addition, the alignment accuracy of global alignment is inferior in principle as compared with die-by-die alignment, but is an accuracy that does not cause a problem in practice. For this reason, an exposure apparatus used in a manufacturing process for mass production of semiconductor devices mainly employs global alignment.

  By the way, in an exposure process for mass-producing semiconductor devices, wafers are processed in units of lots of a predetermined number (for example, 25 or 50) for improving work efficiency. In such lot processing, the exposure apparatus processes all the wafers in the lot according to the processing conditions determined for the first wafer belonging to the lot. However, even though each wafer belonging to the same lot is exposed under the same processing conditions, in an exposure apparatus that employs global alignment, there is a wafer whose pattern superposition of the upper layer pattern layer and the lower layer pattern layer is out of specification. It happens suddenly.

  After the exposure process is completed, the overlay accuracy of the pattern is measured in order to confirm whether or not the pattern has been formed normally. However, since it takes a lot of time to measure the overlay accuracy of the pattern, lot processing measures the overlay accuracy of one wafer belonging to the lot, and if the measurement result is within the standard, the lot All the wafers are considered to be within the standard. In this case, an unexpectedly out-of-standard wafer is not detected at the stage of the exposure process, and a quality defect due to an overlay accuracy defect is detected only in an electrical characteristic inspection performed after the wafer is completed. As described above, a wafer in which a defect is detected after completion cannot be repaired and must be discarded.

For example, Patent Document 1 listed below discloses a technique for detecting a defective lot in such lot processing in an exposure process. In the technique disclosed in Patent Document 1, a predetermined processing error for the wafer is detected while discriminating the type in the exposure process. Then, when the same type of processing error occurs continuously for a wafer in the same lot for a predetermined reference number or more, the processing of the entire lot including the unprocessed wafer is stopped.
JP-A-10-125595

  However, the above-described sudden overlay error also occurs in a wafer that has been normally exposed. That is, a sudden overlay error is not detected as a processing error. For this reason, even when the technique disclosed in Patent Document 1 is applied, since the same type of processing error is not continuously detected, the exposure processing of the lot including the wafer in which the sudden misalignment has occurred is performed. It cannot be canceled. Therefore, the lot including the wafer in which the sudden misalignment proceeds proceeds to the next process, and finally becomes a defective wafer.

  Such sudden misalignment can be detected by measuring the overlay accuracy of the pattern for all wafers after the exposure process. However, it takes a long time to measure the overlay accuracy for all wafers, and the throughput of the manufacturing process is greatly reduced. Such a decrease in throughput increases the manufacturing cost. In order to improve the throughput of the exposure process, the number of alignment mark measurement points may be reduced during alignment. In this case, the number of sudden overlay deviations in the lot is further increased.

  The present invention has been proposed in view of the above-described conventional circumstances, and an exposure system capable of reliably detecting, in an exposure process, a wafer in which a sudden overlay error has occurred without reducing throughput. An object is to provide an exposure method.

  In order to solve the above problems, the present invention employs the following means. First, an exposure system according to the present invention is an exposure system that transfers a pattern formed on a mask onto a substrate by a step-and-repeat method. The exposure system according to the present invention includes an arrangement error measurement unit, an element component calculation unit, a fluctuation range calculation unit, a comparison unit, and a determination unit. The arrangement error measurement unit measures an arrangement error of shots belonging to a specific layer already formed on the substrate. The element component calculation unit decomposes the measured arrangement error into a plurality of element components. The fluctuation range calculation unit calculates a fluctuation range of at least one of the element components obtained for a plurality of substrates. The comparison unit compares the calculated variation width of the element component with a variation allowance preset for the element component, and outputs a comparison result. Then, based on the comparison result output from the comparison unit, the determination unit determines whether there is an abnormality in the exposure process.

  According to this configuration, it is possible to detect a lot including a wafer in which an overlay error has occurred unexpectedly as a lot whose element component fluctuation range is equal to or greater than an allowable width. In addition, each element component used for detecting an abnormal exposure process is data that must be acquired in order to perform alignment in the exposure process. For this reason, it is possible to reliably detect an abnormal exposure process without significantly reducing the throughput of the exposure process.

  For example, an orthogonality component, a scaling component, an offset component, and a rotation component can be employed as the element component. Here, the orthogonality component is a component that causes a positional deviation with a deviation amount corresponding to a straight line passing through the origin of the orthogonal coordinate system with the origin of the orthogonal coordinate system set on the substrate as a reference point. The scaling component is a component in which each shot causes a displacement that linearly expands and contracts with the origin of the orthogonal coordinate system as a reference point. The offset component is a component that causes a displacement in which each shot is translated from the origin of the orthogonal coordinate system as a reference point. Further, the rotation component is a component that causes a displacement in which each shot is rotated around the origin of the orthogonal coordinate system.

  On the other hand, from another viewpoint, the present invention can also provide an exposure method applicable to the above-described exposure system and the like. That is, in the exposure method according to the present invention, first, an arrangement error of shots belonging to a specific layer already formed on a plurality of substrates belonging to the same lot is measured for each substrate. Next, based on each measured arrangement error, a plurality of element components constituting the arrangement error are calculated for each substrate. Furthermore, the fluctuation range is calculated based on at least one element component calculated for each substrate. The calculated variation width of the element component is compared with a variation allowable width set in advance for the element component, and the presence / absence of an exposure process abnormality is determined based on the comparison result.

  Note that an orthogonality component, a scaling component, an offset component, and a rotation component can be employed as the element component. In this case, it may be determined that there is an abnormality in the exposure process when the variation of the orthogonality component and the variation width of the scaling component are each greater than or equal to the allowable variation range. Further, when it is determined that the exposure process is abnormal, the exposure process of the lot can be stopped, or the exposure process determined to be the exposure process abnormality can be performed again on the lot. .

  According to the present invention, it is possible to detect a lot including a wafer in which a sudden overlay error has occurred without measuring overlay accuracy for all wafers. Therefore, it is possible to reliably detect an exposure process abnormality without significantly reducing the throughput.

  Further, since the exposure process abnormality can be reliably detected in the exposure process, the exposure process abnormality can be eliminated by reprocessing the lot. As a result, it is possible to regenerate a wafer that has been discarded as a wafer with poor characteristics in the past, and to improve the non-defective product rate.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the following embodiments, the present invention is embodied as an exposure system including a step-and-repeat type exposure apparatus that employs global alignment.

  FIG. 1 is a schematic block diagram showing the exposure apparatus system of this embodiment. As shown in FIG. 1, an exposure system 1 according to the present embodiment includes an exposure apparatus 2 that performs exposure for transferring a pattern onto a semiconductor substrate (hereinafter referred to as a wafer). The exposure apparatus 2 performs shots belonging to a pattern layer (hereinafter referred to as an upper layer) to be exposed to a specific pattern layer (hereinafter referred to as a lower layer) already formed on the wafer by global alignment. Perform alignment.

  When performing alignment, first, the alignment error measurement unit 21 of the exposure apparatus 2 measures the coordinates of alignment marks formed on a plurality of specific shots belonging to the lower layer. Based on the coordinates, the arrangement error measurement unit 21 calculates a displacement (hereinafter referred to as arrangement error) between the shot arrangement of the lower layer and the shot arrangement of the lower layer. The coordinate measurement of the alignment mark is performed by a known method using a laser or the like.

  The arrangement error is decomposed into simple and regular element components as described below in the element component calculation unit 22 of the exposure apparatus 2. FIG. 2A to FIG. 2D are diagrams illustrating an example of element components calculated by the element component calculation unit 22. In the present embodiment, the arrangement error includes the orthogonality component shown in FIG. 2A, the scaling component shown in FIG. 2B, the offset component shown in FIG. 2C, and the rotation shown in FIG. It is decomposed into a linear component composed of components and non-illustrated nonlinear component elements. 2A to 2D, the design array 10 based on the notch 16 of the wafer 14 is indicated by a solid line, and the shot array 12 in which the displacement of each element component is generated is indicated by a broken line. Yes. In addition, an arrow 13 in FIGS. 2A to 2D indicates the direction of displacement of the shot array 12 with respect to the design array 10.

  As shown in FIG. 2A, the orthogonality component is shifted according to a straight line L passing through the origin of the orthogonal coordinate system with the origin of the orthogonal coordinate system where each shot is set on the wafer 14 as a reference point. It is a component that causes misalignment in quantity. As shown in FIG. 2B, the scaling component is a component that causes a displacement in which each shot is linearly expanded and contracted with the origin of the orthogonal coordinate system as a reference point. As shown in FIG. 2C, the offset component is a component that causes a displacement in which each shot is translated with the origin of the orthogonal coordinate system as a reference point. As shown in FIG. 2D, the rotation component is a component that causes a displacement in which each shot is rotated around the origin of the orthogonal coordinate system. Components that do not belong to any of the linear components shown in FIGS. 2A to 2D are nonlinear components. The nonlinear component is not shown because it is a second-order or higher-order shift component.

  Such an alignment error is caused by wafer distortion after exposure and individual differences in exposure apparatuses (difference in alignment characteristics between exposure apparatuses of the same structure and same model). The distortion of the wafer is caused by, for example, the expansion of the wafer when a heat treatment or the like is performed after the exposure, or warpage of the wafer due to various film stresses formed on the wafer.

  Each element component described above is input to the exposure control unit 24 that controls the drive of the stage on which the wafer is placed in the exposure apparatus 2, and when the upper layer shot is exposed on the wafer 14, the element component (mainly, mainly) Position correction based on the above four types of linear components) is performed. Thereby, the exposure apparatus 2 can perform exposure in which the lower layer and the upper layer are superimposed with high accuracy.

  Further, in the exposure system 1 of the present embodiment, the exposure apparatus 2 includes a fluctuation range calculation unit 23. As will be described in detail later, the fluctuation range calculation unit 23 calculates the in-lot range for at least one element component of the acquired element components of the arrangement error. Here, the in-lot range is a difference between the maximum value and the minimum value of each element component calculated for wafers belonging to the same lot. The in-lot range output from the fluctuation range calculation unit 23 is input to the comparison unit 3. Then, based on the signal output from the comparison unit 3, the determination unit 4 determines whether or not there is a wafer in which the sudden overlay deviation has occurred in the lot.

  FIG. 3 is a flowchart showing processing of one lot in the exposure system 1 of the present embodiment. The exposure system 1 having the above configuration performs exposure after aligning shots of the upper layer with respect to the lower layer by the above-described method when processing a lot of wafer groups. At this time, the processing conditions (the formation conditions of the photoresist formed on the wafer, the exposure time, etc.) for each wafer belonging to the same lot are the same, but the alignment for each wafer is performed for each wafer.

  When performing exposure of each wafer, the arrangement error measuring unit 21 first measures the arrangement error of the lower layer based on the coordinates of the alignment marks formed in a plurality of specific shots (S1 in FIG. 3). The element component calculation unit 22 decomposes the arrangement error into the above five types of element components, that is, an orthogonality component, a scaling component, an offset component, a rotation component, and a nonlinear component (S2 in FIG. 3). Then, the exposure control unit 24 performs exposure according to the shot arrangement of the lower layer by correcting the shot position of the upper layer by an amount corresponding to the calculated element component (S3 in FIG. 3). When the exposure is completed, the exposed wafer is unloaded from the exposure apparatus 2, and the next wafer belonging to the same lot is loaded into the exposure apparatus 2 (S4 Yes in FIG. 3). Arrangement error measurement, element component calculation, and exposure processing are also performed on the next wafer that is carried in.

  When the exposure processing is completed for all the wafers belonging to the lot, the fluctuation range calculation unit 23 calculates the in-lot range of the component component calculated by the component component calculation unit 22 (S4 No → S5 in FIG. 3). Although not particularly limited, in the exposure system 1 of the present embodiment, the fluctuation range calculation unit 23 calculates only the in-lot range Ro of the orthogonality component and the in-lot range Rs of the scaling component. The calculated intra-orthogonality component lot range Ro is input to the orthogonality component comparison unit 31 of the comparison unit 3, and the scaling component intra-lot range Rs is input to the scaling component comparison unit 32.

  When the orthogonality component in-lot range Ro is input, the orthogonality component comparison unit 31 reads the standard value Ra of the orthogonality component in-lot range from the storage unit 33 configured by a hard disk or the like. The orthogonality component comparison unit 31 compares the orthogonality component in-lot range Ro with the standard value Ra read from the storage unit 33 (S6 in FIG. 3). When the orthogonality component in-lot range Ro is equal to or greater than the standard value Ra, the orthogonality component comparison unit 31 outputs a detection signal Sa that can be recognized by the determination unit 4 such as a logical signal “1”, for example (FIG. 3 S6 Yes).

  Similarly, when the scaling component lot range Rs is input, the scaling component comparison unit 32 reads the standard value Rb of the scaling component lot range from the storage unit 33. The scaling component comparison unit 32 compares the in-scaling lot range Rs with the standard value Rb read from the storage unit 33 (S7 in FIG. 3). When the in-scaling component lot range Rs is greater than or equal to the standard value Rb, the scaling component comparison unit 32 outputs a detection signal Sb that can be recognized by the determination unit 4 such as a logical signal “1”, for example (S7 Yes in FIG. 3). ).

  Note that the standard values Ra and Rb are set in the storage unit 33 in advance as predetermined tolerance values that do not cause the semiconductor device to become defective due to the overlay deviation between the upper layer pattern and the lower layer pattern. The standard values Ra and Rb can be determined based on, for example, the values of the element components obtained for a large number of lots.

  As shown in FIG. 1, output signals from the orthogonality component range comparison unit 31 and the scaling component comparison unit 32 are input to the determination unit 4. The determination unit 4 detects whether or not the detection signals Sa and Sb are output simultaneously. When the detection signals Sa and Sb are output at the same time, the determination unit 4 determines that an overlay misalignment wafer exists in the lot (exposure processing abnormality), and the exposure apparatus 2 recognizes, for example, the logical signal “1”. A possible determination signal J is output (S8 in FIG. 3).

  The output of the determination unit 4 is input to the exposure control unit 24 of the exposure apparatus 2. When the determination signal J is input, the exposure control unit 24 reprocesses all wafers belonging to the lot. That is, the exposure process is repeated from the beginning. In this case, the exposed photoresist on all the wafers belonging to the lot is removed, and the photoresist is applied again on all the wafers.

  When the detection signal Sa is not output (S6 No in FIG. 3) and when only the detection signal Sb is output (S7 No in FIG. 3), the determination unit 4 does not output the determination signal J. In this case, the exposure process is terminated and the lot is advanced to the next process. Needless to say, the case where the detection signal Sa is not output (S6 No in FIG. 3) includes the case where the detection signal Sb is not output.

  In the above configuration, the element component calculation unit 22, the fluctuation range calculation unit 23, the comparison unit 3 (the orthogonality component comparison unit 31 and the scaling component comparison unit 32), and the determination unit 4 include, for example, a dedicated arithmetic circuit, It can be realized as hardware including a processor and a memory such as a RAM and a ROM, and software stored in the memory and operating on the processor.

  Hereinafter, the reason why sudden overlay deviation can be detected with the above configuration will be described in detail.

  When the present inventors examined the cause of sudden misalignment, the main factor was found to be mismeasurement of the array error. The erroneous measurement is that the arrangement error measurement means 21 measures the arrangement error of the shots of the lower layer abnormally large or abnormally small. That is, when the shots belonging to the upper layer are aligned based on the erroneously measured arrangement error, the exposure apparatus 2 performs the alignment of the upper layer shots according to the erroneous lower layer shot arrangement. For this reason, overlay deviation occurs between the lower layer and the upper layer. Therefore, it is possible to detect a sudden overlay error by detecting an erroneous measurement of the arrangement error of the lower layer.

  Further, as described above, each wafer belonging to the same lot is processed under the same processing conditions in each process including the exposure process. For this reason, wafer distortion or the like generated during the manufacturing process occurs in a state having a certain tendency in each wafer belonging to the same lot. Therefore, the arrangement error of the shot arrangement of the lower layer measured by the exposure apparatus 2 has a certain tendency with respect to the wafers belonging to the same lot. That is, a wafer in which an arrangement error that deviates from the arrangement error of other wafers belonging to the same lot is very likely to be an erroneous measurement. As described above, by monitoring the range within the lot which is the fluctuation range of the array error of wafers belonging to the same lot and comparing it with a predetermined standard range within the lot, an array error deviating from the array error of other wafers can be obtained. Can be detected. Therefore, by monitoring the in-lot range of each element component, it is possible to detect a wafer in which an unexpected overlay error has occurred. From this point of view, in the in-lot range of the above five types of element components, when at least one of the ranges exceeds the standard value set for each element component, a wafer that has suddenly shifted in overlay is in the lot. It may be determined that it exists.

  However, it is complicated to set standard values for the in-lot ranges of the five types of element components. Therefore, the inventors of the present application have experimentally studied a variety of judgment criteria and a sudden overlay error that has occurred in a large number of actually manufactured lots. I found it. That is, there is a specific element component in which the fluctuation of the value when the erroneous measurement occurs is particularly noticeable. Accordingly, by monitoring only the in-lot range of the specific element component among the five types of element components, it is possible to detect a sudden overlay error. As a result of experimental studies by the inventors, it was confirmed that the orthogonality component and the scaling component fluctuate particularly sensitively. Further, in the present embodiment, when both the in-lot range of the orthogonality component and the in-lot range of the scaling component are equal to or greater than the standard value, it is determined that there is a wafer in which a sudden overlay error occurs in the lot. Yes. As described above, by using the determination criterion that both of the element components, not one of the element components, are equal to or greater than the standard value, it is possible to more reliably detect the sudden overlay deviation. As described above, in the present embodiment, although only the in-lot range of two types of element components is monitored, it is possible to detect a sudden overlay shift with a higher probability.

  FIG. 4 is a diagram showing the in-lot range of the orthogonality component and the in-lot range of the scaling component in a lot (25 wafers / lot) that does not include a wafer in which overlay deviation has occurred. FIG. 5 is a diagram showing the in-lot range of the orthogonality component and the in-lot range of the scaling component in a lot including a wafer in which overlay deviation has occurred. 4 (a) and 5 (a), the vertical axis corresponds to the orthogonality component of the arrangement error, and the horizontal axis corresponds to the wafer number in the lot. 4B and 5B, the vertical axis corresponds to the scaling component of the arrangement error, and the horizontal axis corresponds to the wafer number in the lot.

  4 and 5, the orthogonality component is the angle 17 (clockwise direction) formed by the Y axis of the orthogonal coordinate system set on the wafer 14 and the straight line L, as shown in FIG. Is positive, unit: μrad). The scaling component is indicated by the expansion / contraction ratio (the enlargement direction is positive, unit: ppm) based on the length of one side of the shot in the design array 10. 4B and 5B, the expansion / contraction ratio in the X-axis direction of the orthogonal coordinate system set on the wafer 14 is indicated by a solid line, and the expansion / contraction ratio in the Y-axis direction is indicated by a broken line.

  4 and 5, in the lot (FIG. 4) that does not include the wafer in which the overlay deviation occurred, the in-lot range Ro of the orthogonality component and the in-lot range Rs of the scaling component are very small, that is, between wafers. It can be understood that the variation of the orthogonality component and the scaling component is small. In the example of FIG. 4, the range Ro within the orthogonality component lot is 0.58 μrad. Further, the scaling component in-lot range Rs is 0.26 ppm in the X-axis direction and 0.29 ppm in the Y-axis direction. When exposing each wafer belonging to the lot, the exposure apparatus 2 automatically corrects based on the orthogonality component and the scaling component shown in FIG. 4 and the offset component and the rotation component acquired at the same time. Is going. When the overlay accuracy of all the wafers in the lot was measured, the overlay accuracy of all the wafers was within the standard (no sudden misalignment).

  On the other hand, in the lot (FIG. 5) including the wafer in which the overlay deviation has occurred, it can be understood that the in-lot range Ro of the orthogonality component and the in-lot range Rs of the scaling component are relatively large. In the example of FIG. 5, the orthogonality component in-lot range Ro is 1.78 μrad. Further, the scaling component in-lot range Rs is 0.67 ppm in the X-axis direction and 1.15 ppm in the Y-axis direction. When exposing each wafer belonging to the lot, the exposure apparatus 2 automatically corrects based on the orthogonality component and the scaling component shown in FIG. 5 and the offset component and the rotation component acquired at the same time. Is going. When the overlay accuracy of all the wafers in the lot was measured, the overlay accuracy of the wafer No. 8 was out of specification (there was a sudden overlay error).

  As described above, in the wafer in the same lot, when the shot arrangement of the lower layer is normally acquired, the orthogonality component in-lot range Ro and the scaling component in-lot range Rs are stabilized as shown in FIG. However, when the shot arrangement of the lower layer is suddenly erroneously acquired for some reason such as hardware of the exposure apparatus 2, the orthogonality component in-lot range Ro and the scaling component in-lot range Rs become larger than usual. Therefore, by setting standard values for the orthogonality component in-lot range Ro and the scaling component in-lot range Rs, it is possible to detect a lot including a wafer in which an overlay error has occurred suddenly in the exposure process. it can. Since the standard value varies depending on the type of exposure apparatus used, the same value cannot be used for all exposure apparatuses. In the case of the examples shown in FIGS. 4 and 5, since the lot shown in FIG. 4 is an average in-lot range, the standard value Ra of the orthogonality component in-lot range is set to 1 μrad, and the scaling component in-lot range specification By setting the value Rb to 1 ppm, it is possible to detect a lot including a wafer in which an overlay error occurs unexpectedly.

  As described above, according to the exposure system and the exposure method of the present embodiment, it is possible to detect an overlay error between patterns that occurs suddenly in the exposure process. Further, it can be detected by a simple method of evaluating and determining the in-lot range of element components that have been acquired conventionally for alignment by the exposure apparatus. For this reason, it is possible to reliably detect a sudden overlay shift without reducing the throughput of the exposure process.

  Further, according to the present embodiment, since the overlay deviation can be detected in the exposure process, it is possible to perform reprocessing on the wafer in which the overlay deviation has occurred. Further, when a sudden overlay shift occurs in the lot, the exposure control unit 24 of the exposure apparatus 2 stops the exposure apparatus 2 in accordance with the output signal J of the determination unit 4. It becomes possible to carry out and early detection of device troubles becomes possible. In addition, it is possible to quickly prevent a defective lot that has suddenly shifted from being superposed from proceeding to the subsequent steps.

  Furthermore, in this embodiment, since the exposure process abnormality is detected by the in-lot range of the element component, the exposure process is reliably performed by a simpler process than the method of setting the standard value for each element component and detecting the overlay deviation. Abnormalities can be detected.

  The present invention is not limited to the embodiment described above, and various modifications and applications are possible within the scope of the effects of the present invention. For example, in the above description, an example of an exposure system in which the exposure apparatus 2, the comparison unit 3, and the determination unit 4 are separated has been described, but one or both of the comparison unit and the determination unit may be provided in the exposure apparatus. The arrangement is not particularly limited. In the above-described embodiment, the case where the determination is performed by calculating the fluctuation range of the element components when the exposure processing of the wafers in the lot is completed has been described. However, the arrangement error is measured in the exposure processing of each wafer. It is also possible to employ a configuration in which a determination is made by calculating a fluctuation range each time. Further, in the above-described embodiment, the determination unit firstly detects the comparison result of the in-lot range of the orthogonality component, and then detects the comparison result of the in-lot range of the scaling component. Needless to say, they may be reversed or performed simultaneously.

  Further, in the above, as a particularly preferable form, the presence / absence of overlay deviation is determined only by the in-lot range of the orthogonality component and the scaling component, but a combination of other element components may be adopted. In addition, setting a standard value for each element component and making a determination for each element component are not limited.

  The present invention has an effect that a wafer that suddenly shifts in registration within a lot can be detected in the exposure process without reducing the throughput of the exposure process, and is useful as an exposure system and exposure method.

1 is a schematic block diagram showing an exposure system in an embodiment of the present invention. Schematic showing the component elements of array error The flowchart which shows the exposure process in one Embodiment of this invention Diagram showing in-lot data for orthogonality component and scaling component Diagram showing in-lot data for orthogonality component and scaling component

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Exposure system 2 Exposure apparatus 3, 4 Comparison part 5 Judgment part 6 Storage part 14 Wafer 31 Orthogonality component comparison part 32 Scaling component comparison part Ro Intra-lot range Rs of scaling component Ra In-lot range Ra of orthogonality component In-lot range standard value Rb In-lot range standard value of scaling component

Claims (7)

In an exposure system for transferring a pattern formed on a mask onto a substrate by a step-and-repeat method,
Means for measuring an arrangement error of shots belonging to a specific layer already formed on the substrate;
Means for calculating a plurality of element components constituting the arrangement error based on the measured arrangement error;
Means for calculating a fluctuation range of at least one element component calculated for each of a plurality of substrates;
Means for comparing the calculated variation width of the element component with a variation allowable width set in advance for the element component;
Means for determining the presence or absence of an exposure processing abnormality based on the result of the comparison;
An exposure system comprising: The plurality of element components are:
Each shot of the specific layer, with the origin of the orthogonal coordinate system set on the substrate as a reference point, an orthogonality component that causes a displacement in a deviation amount according to a straight line passing through the origin of the orthogonal coordinate system;
Each of the shots of the specific layer has a scaling component that causes a displacement that is linearly expanded and contracted with the origin as a reference point;
Each shot of the specific layer has an offset component that causes a displacement that is translated from the origin as a reference point;
Each shot of the specific layer has a rotation component that causes a displacement that is rotated around the origin, and
The exposure system according to claim 1, comprising: An exposure method applied to an exposure system for transferring a pattern formed on a mask onto a substrate by a step-and-repeat method,
A step of measuring, for each substrate, an arrangement error of shots belonging to a specific layer, already formed on a plurality of substrates belonging to the same lot;
Calculating a plurality of element components constituting the arrangement error for each substrate based on the measured arrangement errors;
Calculating a fluctuation range of at least one element component calculated for each of the substrates;
Comparing the calculated variation width of the element component with a variation allowable width set in advance for the element component;
Determining the presence or absence of an exposure processing abnormality based on the comparison result;
An exposure method comprising: The plurality of element components are:
Each shot of the specific layer, with the origin of the orthogonal coordinate system set on the substrate as a reference point, an orthogonality component that causes a displacement in a deviation amount according to a straight line passing through the origin of the orthogonal coordinate system;
Each of the shots of the specific layer has a scaling component that causes a displacement that is linearly expanded and contracted with the origin as a reference point;
Each shot of the specific layer has an offset component that causes a displacement that is translated from the origin as a reference point;
Each shot of the specific layer has a rotation component that causes a displacement that is rotated around the origin, and
The exposure method according to claim 3 comprising:   5. The exposure according to claim 4, wherein when the variation width of the orthogonality component is equal to or larger than a predetermined variation allowable width and the variation width of the scaling component is equal to or larger than a predetermined variation allowable width, it is determined that the exposure process is abnormal. Method.   The exposure method according to claim 3, wherein when it is determined that the exposure process is abnormal, the exposure process for the lot is stopped. 4. The exposure method according to claim 3, wherein when it is determined that the exposure process is abnormal, the exposure process determined to be abnormal is performed again for the lot.

Images