TW200907596A - Projecting exposure method, alignment method, and projecting exposure apparatus - Google Patents

Abstract

The present invention provides projecting exposure method and exposure apparatus, wherein the exposure method includes the driving platform, the location signal output mechanism, the standard mark inspection mechanism, and the projection exposure apparatus for location determination mechanism to project pattern and image onto exposure substrate; the exposure substrate has at least nine exposure regions provided for pattern and image projection and at least nine inspection regions provided for inspection substrate to be deformed; the standard mark for the exposure substrate presents at least nine standard locations of the exposure regions and also at least nine standard locations of the inspection regions, which includes the following steps: (1) exposure substrate location control step: after the exposure substrate is used as substrate and loaded on platform, sequentially moving the loading platform through the driving platform to position on at least four inspection locations of the exposure substrate; (2) storage step for exposure substrate location: retrieving each location of exposure substrate inspection on loading platform and storing them; (3) calculation step for standard mark position: based on the exposure substrate inspection location, using the standard mark inspection mechanism to inspect the standard mark on exposure region, and then based on inspection result and loading position to calculate the standard mark position of the exposure substrate; (4) calculation step for error parameter value in whole region: according to the standard mark position of the exposure substrate for the whole exposure substrate, using the least square method to calculate all error parameters in the whole region which is characterized by the error with the displacement of the exposure substrate standard mark; (5) calculation step for error parameter value in inspection region: based on at least two locations of the exposure substrate standard mark in each inspection region to calculate all error parameters in inspection region which is characterized by the error with the displacement of the exposure substrate standard mark; (6) calculation step of the linear relationship for error parameter value in inspection region: based on error parameter value in each inspection region, using the least square method to calculate the linear relationship of the error parameter valve in inspection region; (7) calculation step for differential linear relationship: calculating at least first differential value upon the error parameter value of the inspection region between two neighboring inspection regions; and based on the differential value, using the least square method to calculate the differential linear relationship; (8) calculation step for weighted coefficient: based on error parameter value in each inspection region, linear relationship on error parameter value in each inspection region, and accumulated summation of the ranks of differential method with the differential linear relationship to calculate error information of the error parameter values in inspection region. Therefore, according to error information of the error parameter values in inspection region to calculate weighted coefficients, the coefficients indicates: linear relationship according to the error parameter value in inspection region, error of the linear relationship in differential method, and the proportion and magnitude not based on these errors; (9) determination step for exposure substrate location: based on the standard mark location, the whole error parameter values, and the weighted coefficients of the exposure substrate to calculate the target location ready for positioning with loading platform, and to position the loading platform on the targeted location.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a projection exposure method for forming a pattern on an exposure substrate, an alignment method for the projection exposure, and an optical device capable of realizing the same. [Prior Art] Conventionally, an exposure apparatus for a printed circuit board uses an alignment microscope to make a reference mark formed on a reticle and a reference mark formed on an exposure substrate, and then to retreat the alignment microscope. For exposure. Patent Document 1: Japanese Laid-Open Patent Publication No. 2003-316 No. 14 ▲ The projection exposure apparatus described above does not require an interferometer for measuring the position of the stage, so that the cost is low, and it is widely used in exposure of printed circuit boards, 22! When the processing steps of the seesaw are stretched, it is not high-precision: 〇疋 position. Moreover, if there is an error in the movement of the stage, it is also impossible to accurately position the firefly A..^. Moreover, the printed circuit board has a larger nonlinear error than the wafer base, and the linear error correction used by the exposure device for the tatto-day substrate alone is ***, I, 々 A 〃霄^ Alignment, there are cases where sufficient accuracy cannot be obtained. SUMMARY OF THE INVENTION The present invention is based on the above-mentioned matters of ginger blue, ^ μ, 4 ·, t κ, the purpose of which is to provide a shirt exposure method, a car, earth η, w method and shirt exposure The device can position the substrate in the position without using a position measuring device such as an interferometer of the stage, and even for a printed circuit board with a large nonlinear error, the alignment is fast and high-precision, and even in the exposure The regional peripherals 200907596 are labeled and can also utilize good alignment for non-linear multiple overlapping exposures. The printed circuit board with large error is aligned, the same mark is used for alignment, and the whole layer can be performed, and the above object is achieved. The projection exposure method of the present invention is used for the image exposure apparatus, and the exposure is used again.亢 illuminating the patterned image onto the exposed substrate mounted on the loading platform by irradiating the patterned marking with the pattern; the projection exposure apparatus comprises: a driving stage, and positioning the loading platform for loading the substrate to a minimum of 9 a predetermined position; v a position signal output mechanism disposed on the drive stage and outputting a position signal indicating a position of the loading stage; a reference mark detecting mechanism for detecting an exposure substrate reference mark formed on the exposure substrate, Or a reference substrate reference mark formed on the reference substrate; and a π position determining mechanism that determines the position of the loading stage based on the position indicated by the position signal; wherein the exposure substrate 'has a projection for the pattern image At least 9 exposure areas; °H exposure substrate 'haves at least 9 detection areas for detecting substrate deformation; shai exposure The board reference mark ' indicates a reference position of the at least nine exposure areas, and indicates a reference position of the at least nine detection areas; and includes the following steps: 200907596 Exposure substrate position control step in which the exposure substrate is loaded as the substrate In the loading platform, the loading stage is sequentially positioned to move at least four predetermined exposure substrate detecting positions by the driving stage; and the exposure substrate position storing step is performed by detecting the exposure substrate from the position signal. The position of the loading station is stored and stored; ^ a reference mark position calculating step is performed at each of the exposure substrate detecting positions, and the reference mark detecting mechanism detects the exposure area reference mark, and then the loading table is based on the detection result The position of the exposure substrate reference mark is calculated; & the total area error parameter value calculation step is calculated based on the position of the exposure substrate reference mark based on the position of the exposure substrate reference mark using the least square method The error of the displacement of the mark is the error parameter value of the feature; 5 detection The domain error parameter value calculation step is to calculate a detection region error loss value characterized by an error of displacement based on the exposure substrate reference mark in each of the detection regions, based on at least two positions of the exposure substrate reference marks; ' > The detection region error parameter value linear component calculation step is performed by using the least square method to calculate a linear component of the detection region error parameter value according to the detection region error parameter value in each detection region; the differential linear component calculation step is adjacent to The two detection regions calculate a difference between at least 1 |5 of the spur region error parameter value, and a difference linear component is calculated according to the difference ′ using the least square method; the weighting coefficient calculation step is Each of the detection regions calculates the error parameter value of the detection region according to the detection unit 200907596 regional error parameter value, the linear component of the S-detection region error parameter value, and the cumulative sum of the difference order according to the linear component of a difference Error information, and then calculating the weight based on the error information of the error value of the detection area a weighting coefficient for indicating a linear component of the error value of the detection region and an error of the differential linear component and a ratio of an error according to the error, and an exposure substrate position determining step according to the exposure substrate reference The position of the marker, the overall error parameter value, and the weighting factor calculate a target position at which the loading station is to be positioned, and then the loading station is positioned at the target position. By this processing (the processing of the smart whole chip method), for the position block difference of the reference mark, the error of the approximation by the third-order equation and the magnitude of the random error are used as a reference to determine which is the dominant error, according to the result. , can correct the error of the exposure position. Since Λ ' at the time of exposure, the target position for loading the loading table on which the exposure substrate is mounted can be set to a more correct position. In the projection exposure method of the present invention, the detection region parameter is the X direction

At least one of expansion and contraction, expansion and contraction in the x-direction, rotation in the X direction, rotation in the γ direction, shear deformation in the X direction, and shear deformation in the γ direction. In the projection exposure method of the present invention, error information of the detection region error parameter value is calculated for each of the detection regions, based on a cumulative component of the linear component of the detection region error parameter value and the difference order according to the differential linear component. And the standard deviation of the value obtained by subtracting the error value of the detection region; < the target position at which the loading station is positioned is the linear component of the value of the error in the detection region 200907596 and the cumulative sum of the differential linear components The sum and the detection region error parameter value are calculated based on the weighting coefficient. The exposure apparatus of the present invention for realizing the above-described projection exposure method is a projection exposure apparatus that irradiates exposure light onto a patterned reticle to project the pattern image onto an exposure substrate; a loading stage for positioning the loading platform for loading the substrate in at least nine predetermined positions; a position signal outputting mechanism disposed on the driving stage and outputting a position signal indicating a position of the loading station; a reference mark detecting mechanism for Detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate; and a position determining mechanism determining a position of the loading table based on a position indicated by the position signal; The substrate has at least nine exposed regions for projecting the image of the pattern; the exposed substrate has at least nine regions for detecting deformation of the substrate; and the exposed substrate reference mark indicates a reference position of the at least nine exposed regions, and indicates The reference position of the at least 9 detection areas; 5 hai position determining mechanism, including the following institutions The exposure substrate position control mechanism is mounted on the substrate after the exposure substrate is mounted as the substrate, and the loading stage is manually moved to at least four predetermined exposure substrates by the driving stage. a detection position; an exposure substrate position storage mechanism for acquiring and storing the position of the loading table at each of the exposure substrate detection positions from the position signal; and a "reference mark position calculation means" for each of the #光 substrate detection positions, The reference mark detecting means detects the exposure area reference mark, and then based on the detection result and the position of the loading table, the position of the exposure substrate reference mark is output, and the total area error parameter value calculating means is for the exposure substrate as a whole. And calculating, based on the position of the exposure substrate reference mark, a total area error parameter value characterized by an error according to the displacement of the exposure substrate reference mark; and a detection area error parameter value calculation mechanism in each of the detection areas Calculating according to the position of at least two of the exposed substrate reference marks, according to the exposed substrate The error of the displacement of the reference mark is the characteristic detection area error value; the detection area error parameter value linear component calculation means is used in each of the detection areas, and the least square method is used according to the detection area error parameter value: the detection area error parameter is obtained a linear component of the value; a differential linear component calculation means calculates a difference of at least the detection parameter error parameter values for the two adjacent detection regions, and calculates a differential linear component using the least square method according to the difference; The weighting coefficient calculation means calculates the sum of the regional error parameter value, the linear component of the detection region error parameter value, and the cumulative sum of the difference order corresponding to the differential linear component, in the respective detection regions. Error information of the error parameter value, and then calculating a weighting coefficient according to the error information of the error value of the detection area, the weighting coefficient is used to indicate that the linear component of the error parameter value according to the detection region and the error of the differential linear component are not based on The ratio of the magnitude of the error; A substrate position determining means, based on the exposure position of the substrate reference mark, the overall error parameter value, and said weighting coefficients, the means for calculating a target position of the stage positioning and the loading station is positioned at the opposite pole head. The projection exposure apparatus of the present invention, the detection area parameter is an X direction

At least one of expansion and contraction, expansion and contraction in the Y direction, rotation in the x direction, rotation in the Y direction, shear deformation in the X direction, and frying deformation in the γ direction. The projection exposure apparatus of the present invention causes the error information of the detection region error parameter value to be calculated for each of the detection regions based on the linear component of the detection region error parameter value and the accumulation of the difference order according to the differential linear component And the sum of the standard deviations from the error value of the detection area; I# ^ m The target position at which the loading station is positioned is the linear component of the error parameter value in the detection area and the difference is linearized into ', " The sum of the cumulative sums of atJ F Μ ^ ^ v:t and the value of the check domain decision parameter are calculated based on the weighting coefficient. Conventionally, for the mainstream method of the alignment method of the printed circuit board, the alignment method is performed before the exposure is formed in each exposure area, and the exposure position is corrected according to the position information thereof. The alignment method proposed by the present invention is light. Relative to the production capacity, the first use of all the brush board J is first calculated with a minimum square of nearly 12 200907596 and only the linearity is calculated, u, the whole piece of the exposure is continued. The global ahgnment method. However, the scorpion will have a large nonlinear distortion, which is different from the wafer in one board. In the case of insufficient, the accuracy is in addition to the linearity error. The error is also given to the Orthodox &Sense; Known #早之方,是法. 丌 叉 的 的 的 的 的 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧 智慧'Even if the mark is not able to be detected due to the shape of the mark, it must be able to correspond to it. The tendency of the customer's needs for years has also been used to arrange the exposure on the periphery or centerline of the printed circuit board. The domain-independent mark should be able to correspond to the case where the person does not have internal information. In order to cope with this, in the proposal of the present invention, it is arranged near the outer circumference or the center line of the printed circuit board. The mark can also correct the line error of the exposure position to a predetermined nonlinear error, and can also correspond to the alignment method of the mark detection error (undetectable). Furthermore, the mark arrangement used in the present invention has fewer exposure areas. The number of marked points (number of marks) and the stable mark can be set, therefore, it also has the advantage of providing fast and high precision alignment. The alignment method of the present invention transmits the pattern on the reticle through the projection optical system. When the projection is exposed to the rectangular printed circuit board, the position of the mark on the printed circuit board is detected, and the exposure position is determined according to the mark position information, which is characterized by 'including the following steps: a plurality of pieces arranged in parallel with one side of the rectangular substrate Mark the information, to ^ divided into 2 groups corresponding to a pair of opposite sides; 13 200907596 According to the design coordinates of each mark in the group And the detection information of each marked position' to determine the number of equations having a straight line or using the approximated curve information; for each of the two types determined by the two groups, substituting the design coordinate value for exposure to ' Calculating two pieces of position information; and determining the exposure position by using the inner division information of the two sets of mark coordinates in the axial direction orthogonal to the calculated coordinate axis of the exposure coordinate. The alignment method of the present invention is on the reticle When the pattern is projected and exposed to the rectangular printed circuit board through the projection optical system, the position of the mark on the printed circuit board is detected, and the exposure position is determined according to the mark position information, which is characterized by 'including the following steps: the side of the rectangular substrate The plurality of mark information arranged in parallel is divided into at least two groups corresponding to one pair of opposite sides; the coordinate information of each mark in the group and the detection information of each mark position are determined to have a straight line or According to the curve obtained by the approximation, the curve is determined by the group of the 曝 伢 伢 曝 曝光 曝光 曝光 曝光 曝光 曝光 ' ' ' ' The value of the index is based on the magnification and rotation information of the difference zone before the exposure; and the axis of the exposure coordinate is orthogonal to the calculated coordinate axis, and the coordinates of the coordinate mark are opposite of the && _ Decentralize the news to determine the exposure location. In the alignment method of this preparation month, the plurality of labels /, μ, , ·, which are arranged in parallel with the side of the two directions, are arranged in parallel, and at least the side of the division is divided into 7 2 New 1 J knives and 1 pair of A ' opposite each other and the same processing. 14 200907596 The alignment t and _±_ of the present invention are as determined by the curve: the straight line determined in the group is determined by the number of the standard c by the approximation system, and the one time is marked as the case, 彳β knows 3 When it is 3 o'clock, it is 2 - female ten, and ffi _ ^ is more than 4 points. Human type, the standard is the alignment method of the present invention, which is included in the square of the number of the marked points, and t~ + is preceded by an unrelated step, and #·ΐ| Step; the 4 (four) is determined by the line approximation and the corresponding equation is averaged. The first-order curve approximating the one-person or straight line approximation of the side of the opposite phase is approximated. The human or the second method is included in the processing batch step of the printed circuit board, and the system is said to be straightforward... The step of the indignant judgment is that the star line approximation or the curve approximates the value of the coefficient, making the 2nd κ The number of two people and three times decreased. The mark of the subsequent detection mark is smaller than the point of the i-th slice, and the pattern on the exposure cover of the present invention for realizing the above alignment method is transmitted through the projection hurricane each &+; t is a prior projection exposure device, which is characterized in that , including: "brush board drive stage for loading (4) movement and positioning; ^ very heart-planting position signal output mechanism, clock, _ ^ ° is placed in the § Hai drive stage, and the wheel 矣 - side Position signal of the position of the loading station; and the position detecting mechanism of the unlabeled position, the field of the meter, and the position of the mark; and the alignment control mechanism for detecting the printed circuit board, The detecting mechanism obtains the detection information of the position of 200907596 to determine the exposure position, and controls the driving stage according to the determined exposure position and the position signal; the alignment control mechanism comprises: a marking information classification mechanism for The plurality of mark information arranged in parallel with the i side of the rectangular substrate is divided into at least two groups corresponding to each other; the edge phase number determining mechanism 'based on the design coordinate information of each mark in the group, and The inspection position of each mark position is also used to determine the number of the curve information obtained by the straight line or using the approximation; the position of the calculation unit is determined by the two groups. Each of the equations is substituted for the design coordinate value for exposure to calculate two position information; and the two light position determining mechanism is a relative internal division of the two sets of marker coordinates in the axial direction orthogonal to the calculated coordinate axis using the exposure coordinate The exposure device of the present invention for realizing the above alignment method is a projection exposure device for transmitting a pattern on a reticle through a projection light material and a projection (four) light to a rectangular printed circuit board. The utility model comprises: a driving stage for performing movement and positioning of a loading platform for loading the printed circuit board; a position signal outputting mechanism disposed on the driving stage and outputting a position signal indicating a position of the loading platform; a position detecting mechanism for detecting a position formed on the printed circuit board, and an alignment control mechanism, according to the position of the mark detecting mechanism Detecting information to determine an exposure position, and controlling the driving stage according to the exposure position and the position signal determined by 16 200907596; the alignment control mechanism comprises: a marking information classification mechanism, which is arranged in parallel with the edge of the rectangular substrate The plurality of mark information is divided into at least a corresponding number determining mechanism that is opposite to each other, and the determined coordinate information of each mark in the group and the detection information of each mark position are determined to have a straight line or use A formula for approximating the obtained curve information; a magnification and rotation information calculation means for substituting at least 2 = design coordinates of the diagonal direction of the exposure region for exposure for each of the two types determined by the two groups n is calculated from the difference by the magnification of the exposure area and the rotational exposure position determining the two exposure positions of the axis direction orthogonal to the axis of the mechanism. Using the internal division information of the exposure coordinates and the calculated coordinate set coordinates, the two pairs of orthogonal pairs are opposed to each other in the exposure apparatus of the present invention, and are arranged in parallel in a plurality of sides along the direction. The mark information is divided into at least two groups corresponding to the sides of i and the same processing is performed. It seems that the length of the line is determined by the line in the group or by using the nearest line of the main line. The number is marked by the number of marks. The mark is 2 points:", one-time type mark When it is 3 o'clock, it is a 2nd type, and when it is 4 or more, it is a 3rd type. The exposure apparatus of the present invention includes the steps of determining the order to be calculated in the manner of implementing the method for the number of points. 2009 17 200907596 The first-order type of the side opposite to each other, or 2 the step is selected The equation corresponding to ^ obtained by the straight line approximation is averaged! The second or straight line approximates the second-order approximation of the sub-curve. In the exposure apparatus of the present invention, the first sheet of the processing batch of the printed circuit board is subjected to a step of detecting the mark of 3 points or more for each of the groups; and the step of detecting the mark of 3 points or more; Approximate to the value of the approximation coefficient of the second and third times when the approximation of the approximation curve is approximated, so that the number of points detected by the second H β β is smaller than that of the first slice. The projection exposure apparatus of the present invention The exposure light is irradiated onto the patterned reticle sheet (4). The pattern image is projected on the exposure substrate, and includes: driving the σ, and positioning the loading table for loading the substrate in at least four predetermined positions. a position signal output mechanism 'disposed on the driving stage and outputting a position signal indicating a position of the loading stage; a reference mark detecting mechanism for detecting an exposure substrate reference mark formed on the exposure substrate or formed on the reference substrate a reference substrate reference mark; and a position determining mechanism that determines a position of the loading table based on a position indicated by the position signal; the exposure substrate reference mark is used to indicate At least four marks of the reference position of the exposure substrate are formed; and the reference substrate reference mark is composed of at least four marks indicating the reference position of the reference substrate; 18 200907596 The position determining mechanism includes ·· a vertical correction mechanism 'for correcting a position error of the loading stage on the side of moving the loading; and a position: difference VI setting correction mechanism for correcting a linear component error in the position difference of the exposure substrate reference mark; The first position correcting mechanism includes: a reference substrate position control mechanism that, after the reference substrate is mounted on the loading table, moves the loading table in sequence by at least a cough ', " soil plate moving stage Three predetermined reference substrate detection positions '· A standard::: plate position storage mechanism, the position of the loading table at each of the quasi-substrate detection positions is obtained from the position signal and stored; the reference substrate fiducial mark position storage mechanism And detecting the reference substrate reference by the reference mark detecting mechanism at each of the reference substrate=positions, and then according to the detection result thereof The position of the loading platform is calculated, and the position of the reference mark of the reference substrate is calculated and stored; / The reference substrate position correction is performed by the left side of the substrate, and the nasal storage mechanism is based on the position of the reference substrate based on the drought target 5 "P 4 ^士& is used to fork the loading position correction data of the position error of the loading table and store it; the second position correcting mechanism comprises: an exposure substrate position control mechanism, which is exposed on the Du Qiao exposure substrate When the substrate is mounted on the loading platform, the mounting stage is sequentially positioned by the driving stage to position at least three predetermined exposure substrates determined by the loading position correction data. Position; the exposure substrate position occupant mechanism obtains the position detection position of each of the 19 200907596 exposure substrates from the position signal by the (four) exposure substrate reference mark position error #_ sub detection position, by which the reference mark inspection machine == Exposing the substrate mark, and then according to the detection result and the position of the optical substrate reference optical substrate reference mark, and storing (4); calculating the exposure linear deviation correction operation mechanism by mouth and According to the exposure substrate material storing the mark position storage mechanism - the linear error correction resource of the linear component error, the value of the loading table position correction data is used to correct the loading station ΐ: "driver drive control Positioning of the loading table: the level of the formation error of the reference mark of the reference substrate: for measurement, the pattern can be transferred to the exposure = two standard; the long dimension measurement and the averaging effect Moreover, the positioning accuracy of the loading platform can be improved by reducing the number of measuring points of the base material 5, and the manufacturing capacity of the substrate can be improved. The projection exposure apparatus of the present invention comprises: a digital micromirror element having a plurality of mirrors, and wherein a direction of reflection of light incident on the plurality of mirrors is determined by the plurality of mirrors; and the microarray a lens having a plurality of microlenses corresponding to the plurality of mirrors respectively; and a point image formed by the microarray through the fourth image is projected onto the #optical substrate, characterized in that it comprises: a driving stage, which is to be loaded The loading platform of the substrate is positioned at at least 20 predetermined positions of 200907596; a position signal output mechanism is disposed on the driving stage, and outputs a position signal indicating a position of the loading platform; and a reference mark detecting mechanism for detecting the formation Exposing the substrate reference mark of the substrate or the reference substrate reference mark formed on the reference substrate that can be exposed; and the position determining means determining the position of the loading table based on the position indicated by the position signal; The base substrate reference mark is composed of at least three marks indicating a reference position of the exposed substrate; the reference substrate reference mark is at least three marks indicating a reference position of the reference substrate The position determining mechanism includes: a first position correcting mechanism for correcting a position error of the loading table generated when the loading table is moved; and a second position correcting mechanism for correcting the exposure substrate reference mark a linear component error in the position error; the first position correcting mechanism includes: a reference substrate position control mechanism, wherein the reference substrate is mounted on the loading platform as the substrate, and the loading stage is used to mount the loading table And sequentially positioning and positioning at at least three predetermined reference substrate detecting positions; the reference substrate position storing mechanism is configured to obtain the position of the loading platform at each of the reference substrate detecting positions from the position signal and store the same, and the reference substrate exposure mechanism At the detection position of each of the reference substrates, 21 200907596 2 light is irradiated onto the reference reticle to be formed on the reference And the reference substrate forms the 2 quasi-substrate position correction calculation storage mechanism, and the reference mark inspection structure detects the reticle reference mark formed on the reference substrate The reference substrate reference mark calculates the relative position of the image of the reticle reference and the reference mark of the reference substrate based on the detection result, and the two pairs are placed at the position of the loading table to calculate the position of the loading table. 4 position correction data of the position difference is stored and stored; the second position correction mechanism includes: and the position control mechanism, when the exposure substrate is used as the substrate=loading station, by the driving stage, The loading stage sequentially moves at least three predetermined exposure substrate detection positions determined according to the position correction data of the stage; the exposure substrate position storage mechanism obtains the detection position of each cough exposure substrate from the position signal Positioning and storing the loading table; detecting the ϋ base Ϊ reference mark position storage mechanism, each of the exposure substrate = position 4, the reference mark detecting mechanism To detect the exposure substrate reference center 5' and then calculate the exposure substrate according to the detection result, and add the ", and the position of the 3H loading table" to calculate the exposure substrate reference mark. Position and storage; and linear error correction calculation broadcast "Similar machine based on the position of the exposure substrate reference mark stored in the exposure substrate reference storage mechanism, using the most = flat method to calculate the linear component error Linear error correction capital 22 200907596 The projection exposure (four) of the present invention comprises: a digital mirror element; a complex subtractive mirror, and the reflection direction of the light of the (four) (four) complex money mirrors respectively depends on the plurality of reflection errors - · a mirror and a microarray lens having a plurality of microlenses corresponding to the plurality of mirrors respectively corresponding to the occupation #, i / (1); projecting the point image formed by the microarray lens to the sleep main exposure a first substrate, comprising: a driving stage, wherein the loading stage for loading the substrate is moved and positioned in at least three predetermined positions; the position §fl output mechanism is set in 兮μ叙# &amp ; ^ 1 official. a IE moving stage, and outputting a position signal indicating a position of the loading stage; a reference mark detecting mechanism for detecting exposure formed on the exposure substrate, a substrate reference mark, or a reference substrate reference mark formed on the reference substrate And the position determining unit determines the position of the loading platform based on the position indicated by the position signal; and the exposure substrate reference mark is formed by at least three marks indicating a reference position of the exposure substrate; The reference substrate reference mark is composed of at least three marks for indicating a reference position of the reference substrate; the position determining mechanism includes: a first position righting mechanism for correcting when the loading table is moved a position error of the loading stage; and a second position correcting mechanism for correcting a linear component error in a position error of the exposure substrate reference mark; 23 200907596 The first position correcting mechanism includes: a reference substrate position control mechanism, After the reference substrate is mounted on the loading stage as the substrate, the loading stage is mounted by the driving stage. The movement is positioned at at least three predetermined reference substrate detection positions; and the reference substrate position storage device obtains the position of the loading table at each of the reference substrate detection positions from the &signal; and stores the reference substrate reference The mark position storage means detects the reference substrate reference mark by the 6-inch reference mark detecting means at each of the reference substrate positions, and calculates the position of the reference substrate reference mark based on the detection result and the position of the loading stage. And the storage is performed; and the reference substrate position correction operation is performed, and the 隹μ separate storage mechanism calculates the position error of the load port based on the reference substrate reference position. The position correction data of the load port is stored and stored; the second position correction mechanism includes: an exposure substrate position control broadcast spectrum, +β, and is loaded and driven by the money field plate when being loaded on the loading platform as the substrate The stage 'moves the positioning a to the exposure substrate detection position determined according to the loading stage and determined by the Beco; The light substrate position storage mechanism, Λ# „ is obtained from the position signal at the position of the loading table at each of the exposure substrate detection positions and stored; the exposure substrate reference mark position is stored. @捕机构, at each edge Warm also | The detection position, by which the reference mark is picked up to expose the substrate mark, and then according to the detection result 盥 'the position of the reference plate of the earth reference light substrate and the 亥#出5海卞卞M storage; and 24 200907596 P Position=杈-calculation mechanism, based on the position of the exposure substrate reference mark stored in the reference substrate of the exposure substrate, using the most accurate tool for correcting the linear component error of the linear component error The value is used to correct the loading stage's drive control by the driving stage, whereby the positioning of the loading stage can be ordered at the level of the formation error of the reference mark formed on the reference substrate. Therefore, it is not necessary to have a position measuring device such as an interferometer as a position measuring device for the loading table, and it is possible to transfer the pattern to the exposed substrate in a one-piece manner and to have a long-length measurement and averaging effect in a one-piece manner, and by reducing Base: The number of measurement points marked to improve the positioning accuracy of the loading table, i, can also improve the manufacturing capacity of the exposed substrate. - Further, the conductor pattern can be formed on the exposure substrate without using the reticle, and the conductor pattern can be made into a desired shape by the control of the digital micromirror element. Furthermore, by controlling the digital micromirror element, not only the degree of expansion but also the degree of orthogonality can be corrected. The projection exposure apparatus of the present invention includes a projection optical system for irradiating the exposure light to the exposure substrate; the reference mark detection mechanism includes an alignment optical system disposed between the projection optical system and the loading stage; Aligning the optical system, irradiating the non-exposure light to the exposure substrate reference mark or the reference substrate reference mark to detect the exposed substrate reference mark or the reference substrate reference mark; the alignment optical system is detecting the After exposing the substrate reference mark or the 25 200907596 reference substrate reference mark, the substrate is positioned at the detection position; after the detection of the exposure substrate reference mark or the reference substrate reference mark is completed, the position is retracted from the detection position. . By irradiating the non-exposure light to the reference substrate reference mark, the reference mark can be accurately detected. Moreover, in the case of detecting a plurality of reference marks, the alignment optical system is retracted after the final reference mark of the 4» is measured, and then the exposure is started, so that it is not necessary to separate between the plurality of exposure regions. The movement of the alignment optical system shortens the time required for the reference mark detection, thereby increasing productivity. The projection exposure apparatus of the present invention is calculated based on the average value of the positions of the exposure substrate reference marks at the plurality of reference substrate detection positions. Since the loading table position correction data is calculated based on the average value of the positions of the exposure substrate reference marks, an error occurs in the movement of the driving table, or an Abbe error occurs due to the slight deflection of the driving table. Production shape can also further improve the positioning accuracy of the loading platform. In the projection exposure apparatus of the present invention, the reference substrate reference mark is formed on a lattice-like intersection at a predetermined interval in the reference substrate; and the loading position correction data is based on a reference substrate located at the lattice intersection Calculated by the position of the reference mark; the exposure substrate position control means calculates the position of the loading station by using the curve position approximation or interpolation method, and then positions the loading table The target location. When the error gradually changes according to the position of the loading platform, the reference can be shortened by the approximation or interpolation calculation of the position correction data folder of the 200907596 stage. The position measurement time of the interpolated flooding, the storage load port occupied by the correction data is 4, and the time of the mobile drive operation is short. The projection exposure apparatus of the present invention includes: a projection optical system capable of changing a projection projection of the pattern image to the exposure substrate; and determining a projection magnification of the D-hai image, ^ according to the linearity error The linear error correction calculation calculated by the correction computer . is limited to the projection magnification.

Since the linear error correction Shenzu is paid according to the result of the long-size measurement in #4f& E, the average effect can be obtained by using the least squares of the nearest square, and the projection is corrected based on the result. Magnification, . Flatten the pattern to the exposed substrate in the correct size. 1 Further, in particular, depending on the parameters of the linear error correction data, there are, for example, a degree of expansion in the X direction or a degree of expansion in the γ direction. When a projection optical system designed in an asymmetric manner is used, the magnification correction can be independently applied to the rubbing direction and the rubbing direction, and the pattern can be more reliably transferred to the exposure substrate. In the projection exposure apparatus of the present invention, the exposure light is irradiated onto the patterned reticle to project the pattern image on the exposure substrate, and includes: a drive stage 胄 loading platform for loading the substrate Positioning at at least four predetermined positions; a position signal output mechanism disposed on the drive stage and outputting a position signal indicating a position of the loading stage; 27 200907596 a reference mark detecting mechanism for detecting an exposure formed on the exposure substrate a substrate reference mark or a reference substrate reference mark formed on the reference substrate; and a position determining means for determining a position of the loading stage based on a position indicated by the position signal; and a substrate for displaying the image At least 4 exposure areas; a substrate exposure reference mark, an exposure area reference mark for indicating a reference position of the at least four exposure areas; and a position determining mechanism including an exposure position correction mechanism for correcting the exposure substrate Position error of the reference mark; the exposure position correction mechanism includes: = light substrate position control Mechanism, is used as the substrate in the exposed substrate. When the load port is used, the drive stage is driven by the drive stage, and the loading stage is sequentially positioned at at least four predetermined exposure substrate detection positions; the eye position misregistration mechanism is obtained from the position signal at each of the exposure substrates. Positioning the storage station and storing it; the exposure area reference mark position is stored at the detection position, and the reference structure is used to detect the exposure area reference ^' under the exposure substrate, and then according to the detection result Calculating the position of the region reference mark with the loading table position and storing the 丨° position error processing mechanism, and the position of the exposure substrate reference mark is nosed according to the ^1 Μ # M it ^ sr ...j The nonlinear component of the kilo-J position error 28 200907596 Error information; The differential error processing mechanism calculates the exposure using the least squares method based on the difference between the positions of the exposure region reference marks stored in the exposure region reference mark position storage mechanism Error information of nonlinear components in the difference error of the position of the substrate reference mark; and the exposure substrate position The determining means applies weighting according to at least one of the two non-linear components to calculate a target position to be positioned by the loading station, and then positions the loading station at the target position. The error and the error of the nonlinear component are weighted to calculate the target position to be positioned on the loading platform. Therefore, it is possible to determine the proximity of the substrate control according to the state of the exposed substrate. Close to the whole stage of the stage control, and can be used to adjust the positioning accuracy. Here the "weighting" (four) meaning, not only the linear error and linear error & A small to choose the piece by piece mode and the whole piece It is: - square, and can also be emitted according to the magnitude of the nonlinear error and the linearity error in the reticle formed with the illumination pattern formed by the projection exposure apparatus of the present invention to project the pattern image The substrate is exposed to include: three predetermined positions; the driving stage moves the loading platform for loading the substrate to at least the position signal output mechanism, and is disposed at the display a position signal of the position; at the #(four) stage, and an output table reference mark detecting mechanism for detecting a light substrate reference mark formed on the exposure substrate 29 200907596; and a position determining mechanism according to the position indicated by the position signal Determining the position of the loading platform; the exposure substrate 'having at least three exposure regions for projecting the pattern image; and the exposure exposure panel marking for displaying the exposure region reference of the reference position of the at least three exposure regions a position determining mechanism including an exposure position correcting mechanism for correcting a position error of the exposure substrate reference mark; the exposure position correcting mechanism comprising: an exposure substrate position control mechanism, wherein the exposure substrate is mounted as the substrate After the loading stage, the loading stage is sequentially positioned to move at least three predetermined exposure substrate detecting positions by the driving stage; and the exposure substrate position storage mechanism is obtained from the position signal for detecting the exposed substrate. Position of the loading station and (4) deposit; ^ Human exposure area reference mark position storage mechanism, In each of the exposures: the measurement position, the reference mark detection mechanism is used to check: = then according to the detection result and the position of the loading station = the position of the first area reference mark and stored; exposure linear error correction] heavy Zeng Deji The position is stored as “胄, the mechanism, according to the position of the reference area of the exposure area stored in the reference area storage area of the exposure area... The small flat method calculates the position of the use of w ^ soil early, the most used; X positive linear component error The linear error correction factor difference dissipating mechanism is used to calculate the difference between the exposure area reference mark 30 200907596 and the position of the exposure and area reference mark of the position storage mechanism, and the overlapping control mechanism will use the The expansion and contraction obtained by the least square method, and the difference between the orthogonal and the f, are replaced by the difference of 2, and the exposure light is irradiated with the second light: two pieces 'to display and control the image of the pattern on the exposed substrate. The overlapping target position of the exposure substrate. Since at least one of the expansion, the rotation, and the orthogonality obtained by the least square method is substituted into a difference, the processing can be simplified and the productivity can be improved. The projection exposure method of the present invention uses a projection exposure device to illuminate (four) the exposure light into a patterned reticle 'to project the pattern image onto an exposure substrate loaded on a loading stage; the projection exposure apparatus includes: a loading stage for positioning the loading platform for loading the substrate in at least three predetermined positions; a position signal outputting mechanism disposed on the driving stage and outputting a position signal indicating a position of the loading station; a reference mark detecting mechanism a reference substrate reference mark formed on the exposed substrate or a reference substrate reference mark formed on the reference substrate that can be exposed; and a position determining mechanism that determines the loading stage based on the position indicated by the position signal The position of the exposure substrate is defined by at least three marks indicating the position of the base 31 200907596 of the exposed substrate; the reference mark of the reference substrate is used by the company. It is composed of シ3 marks indicating the position of the base sheep of the reference substrate; and includes the following steps: a position control step in which the reference substrate is loaded on the loading table, and the loading table is privately clamped to at least three predetermined reference substrate detection positions by a mourning~dreaming; The substrate position (4) storing step is to obtain the position of the cymbal stage at each of the reference sulcus measurement positions from the position signal and to store the reference substrate exposure step, and to perform the exposure light at each of the reference substrate detection positions Irradiating the reference reticle, and projecting the standard US:: formed on the reference reticle to the reference substrate to form the image of the "line reference mark" on the reference substrate; and the reference substrate position correction calculation storage step The reference mark detecting means detects an image of the reticle I mark formed on the reference substrate and the reference substrate reference mark, and calculates an image of the reticle reference mark and the reference mark of the reference substrate based on the detection result. Position, and then calculating and loading the loading table position correction data for correcting the position error of the loading table according to the continued (four) position and the position of the loading station; The control step is such that after the exposure substrate is mounted on the loading platform as the substrate, the loading stage is sequentially moved by the driving stage to at least three predetermined levels determined according to the loading position correction data. The exposed substrate position detecting step is obtained by taking the position signal from each position of the 32 200907596 exposure substrate detecting position q and storing it. Exposing the substrate reference mark position, detecting the position, by the reference mark f detects the exposure in each of the first substrate inspections, and then extracts the exposure I and T according to the position t of the substrate reference substrate reference mark; and the linear error correction operation + In the step of determining the position of the exposure substrate reference mark stored in the exposure substrate reference mark position storage step, the method for calculating the material is calculated using the least square method. A linear error correction projection projection apparatus having only a positive linear component error, and exposing the pattern image to the projection exposure method of the present invention, using the light to be irradiated onto the patterned reticle, which is carried on the loading platform Exposing the substrate; the projection exposure comprises: driving the stage 'moving the loading table for loading the substrate in at least three predetermined positions; the position signal output mechanism, disposed on the driving stage, and outputting the loading table Position signal; > reference mark detecting means for detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate; and a position determining means according to the position signal Positioning to determine the position of the loading platform; wherein: the exposure substrate reference mark is formed by at least three marks indicating a position of the base 33 200907596 of the exposure substrate; the reference substrate reference mark is Forming at least three marks indicating the reference position of the reference substrate; and including the following steps : a reference substrate position control step of sequentially positioning the loading stage by at least three predetermined reference substrate detection positions by the drive stage after the reference substrate is mounted on the D-loading stage as the substrate; The reference substrate position storing step is to acquire and store the position of the agricultural stage at each of the reference substrate detection positions from the position signal; and the reference substrate reference mark position storing step is performed on each of the reference substrates: the measurement position & the reference mark detecting means detects the reference substrate reference mark, and then the disk m according to the detection result thereof. Wo and. The position of the base load table is calculated and stored in the reference position of the base soap substrate; the reference substrate position correction i is placed on the base + μ W storage step, based on the position of the reference substrate base The error loading position D position is positively stored and stored; the exposure substrate position control step is performed after the substrate is mounted on the I stage as the substrate, and the loading stage is driven by the ^ Depending on the movement positioning, at least three predetermined exposure substrate detection positions are extracted according to the loading position of the loading table; y 3 is similar to the exposure substrate position storage step, and the position detection signal is obtained from the position signal. The loading is performed at each position of the 戟σ and stored; the exposure substrate reference mark position is less vacant, and is performed at each of the continuous umbrella detection positions, and the reference mark is used to detect the "+ first substrate mechanism Detecting the exposure substrate reference 34 200907596 mark 'and then according to the position of the detection junction light; ^ very ^ ° 装 ^ ^, calculate the position of the exposure substrate reference mark and store it and position = = slave operation step 'According to the method of taking the (4) filament reference mark stored in the exposure substrate reference mark 0 (4), 屮ffl r, > τ ^ material. The linear error correction for the positive linear component error is corrected according to the above-mentioned loading and closing, 摅... the value of the position correction data to correct the loading station according to the "driving drive control", thereby being able to form the reference The substrate is Α 挪 挪 ^ ^ ^ ^ , , , , , , , , , , , , ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成 形成The transfer to the exposure substrate and the long-size measurement and averaging effect of the film method can improve the positioning accuracy of the loading table by reducing the number of measurement points of the base mark, and can also expose the manufacturing capacity of the substrate. The projection exposure method of the present invention uses a projection exposure apparatus micro-array lens to form a point, and the sturdy, sturdy stalks from the point of view to the exposure substrate loaded on the loading platform; the projection exposure apparatus comprises: The stage 'will be used to load the substrate, and the mobile station is positioned at at least three predetermined positions; the position signal output mechanism is disposed on the driving stage, and the position of the output table is outputted. a reference mark detecting mechanism for detecting an exposure substrate reference mark formed on the exposure substrate or a reference base 35 200907596 plate reference mark formed on the reference substrate for exposure; a position determining mechanism according to the position signal Positioning to determine the position of the loading station; the digital micromirror element having a plurality of mirrors and enabling the direction of reflection of light incident on the plurality of mirrors to be dependent on the plurality of mirrors; and the microarray lens And a plurality of microlenses corresponding to the plurality of mirrors respectively; wherein: the a-substrate substrate reference mark is formed by at least three marks indicating a reference position of the exposed substrate; the reference substrate reference The mark is composed of at least three marks indicating a reference position of the reference substrate, and includes the following steps: a reference substrate position control step in which the reference substrate is mounted as the substrate on the loading stage by the drive Carrying the stage, the loading stage is sequentially moved and clamped to y 3 predetermined reference substrate detecting positions; The quasi-substrate position storing step acquires and stores the position of the loading table at each of the reference substrate detecting positions from the position signal, and the reference substrate exposing step irradiates the exposure light to the reference at each of the reference substrate detecting positions. a reticle for projecting a reticle reference mark formed on the reference reticle on the reference substrate to form an image of the reticle > 5 fiducial mark on the reference substrate; and a reference substrate position correction operation storing step The reference mark detecting unit 36 200907596 detects an image of the reticle reference mark formed on the reference substrate and the reference substrate reference mark, and calculates an image of the reticle reference mark and the reference substrate reference based on the detection result. Corresponding position of the mark, and then calculating and storing the loading table position correction data for correcting the position difference of the loading table according to the position of the phase: position_stage; the exposure substrate position control step is performed on the exposure substrate After the substrate is loaded on the loading platform, # is driven by the driving stage, and the loading table is sequentially moved according to the loading. At least three predetermined exposure substrate detection positions determined by the position correction data; and the exposure substrate position storage step is to acquire and store the position of the loading table at each of the exposure substrate detection positions from the position signal; and expose the substrate reference mark The position storage step is to detect the exposure substrate reference I by the reference mark detecting mechanism at each of the measurement positions, and then calculate the exposure substrate reference mark based on the detection result and the position of the loading table. Position and storage; and 'according to the position of the reference mark of the light substrate at the exposure substrate reference mark, the linear error correction linear error correction linear error correction operation step stored in the position storage step is used to calculate The information. Microplate The dot image formed by the array lens of the projection exposure method of the present invention, 'Using a projection exposure apparatus, the projection exposure apparatus projected by the exposure stage mounted on the loading stage includes: a driving stage, which is used for the rhyme The loading platform of the 妃+ 褒-loading substrate is moved to the position of 37 200907596 and has three predetermined positions; the position §fl output mechanism is disposed on the driving stage, and outputs a position signal indicating the position of the s-shaped loading platform; a reference mark detecting means for detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate; and the position determining means 'determining the position of the loading stage based on the position indicated by the position signal a digital micromirror element having a plurality of mirrors, and wherein a direction of reflection of light incident on the plurality of mirrors is dependent on the plurality of mirrors; and the microarray lens has a plurality of reflections Mirrors respectively corresponding to a plurality of microlenses; characterized in that: a remotely exposed substrate reference mark is used to represent the exposed substrate The reference substrate reference mark is composed of at least three marks for indicating the reference position of the reference substrate by using _ field field; and includes the following steps: : quasi-substrate position control Step: the reference substrate is used as the substrate: the loading stage is mounted on the loading stage, and the loading stage is sequentially moved and positioned at at least three predetermined reference substrate detecting positions; the reference substrate position storing step , choose you > u. From the location signal, get in each

The reference substrate detects the position of the load A at the position of the battle and stores it; 38 200907596 The check mark = mark position storage step is performed on each of the reference substrates, and the TM mark check mechanism detects the base substrate base The position of the reference mark of the sheep substrate is stored; the reference substrate position correction operation is stored in the phase of the reference V, and the position of the load port of the position error of the loading table is based on the reference substrate base ^^W^τ -^ The right substrate is stored and stored; the exposure substrate position control step is performed after the substrate is mounted on the loading platform as the substrate, by the pole ^ ^ ^ , . . , p ^ 'moving port ' The loading station sequentially moves the clamp position at at least three predetermined exposure substrate detection positions determined according to the position of the loading station; the exposure substrate position storage step, the 俜"攸" Hai position signal is obtained in each The exposure substrate detects the position of the clothing 厥 of the loading a and stores it; the exposure substrate reference mark position stores the mouse, and the direct storage step is performed at each of the exposed substrate detection positions by the reference mark The detection unit 2 detects the exposure substrate reference mark, and then calculates and stores the position of the optical substrate reference mark based on the detection result and the position of the loading stage; and &quot;, linear error correction operation step, & 4 ++ & θ 异 ^, according to the position of the exposed substrate reference mark stored in the exposure substrate reference mark position storage step, using the least square method to calculate 线性 to correct the linear component (4) linear error correction. The value of the loading table position correction data is used to correct the loading position, and the driving control of the driving table is implemented, whereby the positioning of the loading table can be formed in the shape of the I-standard substrate of the I-standard substrate. Come in 39 200907596 仃. Therefore, there is no need to have an interferometer or other position measuring device as the position measurement square cut (4) printed area plate, and the long-size measurement and averaging effect of the slice method, and can be reduced by the reference The number of measuring points is used to improve the positioning accuracy of the loading table, and the manufacturing capacity of the exposed substrate can also be improved. - The conductor can be taken without using the reticle The case is formed on the exposure substrate, and the control of the digital micro-mirror element enables the conductor pattern to have a desired shape. Furthermore, by controlling the digital micro-mirror element, it is not only limited to the degree of expansion, but even orthogonal. The projection exposure apparatus of the present invention, wherein the projection exposure apparatus includes a projection optical system for irradiating the exposure light to the exposure substrate; the reference mark detection mechanism includes a projection optical system The disk μ irradiates the exposure substrate reference mark or the reference substrate reference mark to the exposure substrate reference mark or the reference substrate reference mark; and detects the reference substrate reference mark by the reference mark detection mechanism Or when the substrate reference mark is exposed, the method includes the steps of: positioning the alignment optical system at the detection position when the exposure substrate reference mark or the reference substrate reference mark is not detected; and ending the exposure substrate reference mark or After the detection of the reference substrate fiducial mark, 'position the alignment optical system from the inspection The position where the position is retracted is measured. The reference mark is detected by the non-exposure light. Moreover, when the reference mark is placed on the reference substrate, and it is possible to reliably detect a plurality of reference marks, 200907596 is after detecting the position of the last reference mark, and then the alignment is separated from '' and then the exposure is started. Therefore, it is not necessary to The movement of the alignment optical system between the plurality of exposures can shorten the detection of the reference standard and thus increase the productivity. "In the meantime, the position of the loading table of the projection exposure method of the invention is calculated based on the average value of the exposure of the plurality of reference substrate detection positions. The position correction data is calculated by calculating the average value of the Him position. The error occurs when the drive table moves, or the micro tilt of the drive table is generated, and the positioning accuracy of the loading table can be improved step by step. In the case of (4), the projection exposure method of the present invention is still in the form of a lattice-like intersection in the plate at the predetermined interval in the U.S. base plate; ...^ The position of the reference substrate fiducial mark is calculated at the grid-like intersection and includes the following steps: Using the pre-installation a # ® &gt; Fashen-san is the right data and uses curve approximation or interpolation To position the loading table, ^ ^ is at the target position. After the setting, the loading stage will be used as the error system corresponding to the loading a. When the loading station position correction data changes, because the data is inserted, dream this At &quot Line approximation or interpolation calculation Obtaining the position of the internal stage = capital: = position measurement time of the reference mark, reducing the storage capacity of the door during the calculation performed before the stage mounting, and shortening the moving exposure 41 200907596 The projection exposure method of the present invention includes a projection optical system, The projection magnification of the pattern image may be changed to project the pattern onto the exposure substrate; and the method includes the step of: determining the projection magnification based on the linear error correction data calculated by the linear error correction operation step. It is obtained by the result of the long dimension measurement, and therefore, the averaging effect of the least square approximation calculation can be obtained, and according to the result, the projection magnification can be corrected, and the pattern can be transferred to the exposure substrate in the correct size. According to the parameters of the linear error correction data, for example, the degree of expansion in the X direction or the degree of expansion in the γ direction. When the projection optical system designed in an asymmetric manner is used, the magnification correction can be performed independently for the χ direction and the γ direction. The pattern can be transferred to the exposure substrate more accurately. The projection exposure method of the present invention Using a projection exposure device, the exposure light is irradiated onto the patterned reticle to project the pattern image onto the exposure substrate loaded on the loading platform; the projection exposure device comprises: a driving stage, which is to be loaded The loading platform of the substrate is positioned and positioned in at least four predetermined positions; the position and number output mechanism is disposed on the driving stage, and outputs a position signal indicating the position of the loading station; &quot;reference 'β has been detected by the machine Detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate; and 42 200907596 a position determining mechanism that determines a position of the 裴 stage based on the position indicated by the position signal; The exposure substrate has at least four exposure regions for projecting the pattern image; an exposure region reference mark indicating a reference position of the at least four exposure regions for the exposure substrate reference mark m; and the following steps: exposing the substrate position a control step of mounting the exposure substrate on the loading table as the substrate, by the driving a loading stage for sequentially positioning the loading table at at least four predetermined exposure substrate detecting positions; and an exposure substrate position storing step of obtaining the position of the loading platform at each of the exposed substrate detecting positions from the position signal and storing &quot; an exposure area reference mark position storing step of detecting the exposure area reference mark at each of the exposure substrate detection positions, and then calculating the exposure area based on the detection result and the position of the loading stage; The position of the reference mark is stored; the position error processing step calculates the nonlinearity in the position error of the exposure substrate reference mark using the least square method according to the position of the exposure area reference mark stored in the exposure area reference mark position storage step The error information of the component; the differential error processing step calculates the difference error of the position of the exposure region reference mark using the least square method according to the difference between the positions of the exposure region reference marks stored in the exposure region reference mark position storing step 43 200907596 Medium nonlinear The error information; and a step of exposing a substrate is positioned, (iv) based on the two of the non - error information to the weighting applied 'to the target position is calculated to the placing table' and then placing table in the target position. Since the error of the linear component and the error of the nonlinear component are weighted to calculate the target bit 4 to be positioned on the loading platform, @this, the state of the optical substrate, to determine the bean 夕 夕 w w w ” ”载 Among them, the stage control close to the piece-by-chip method is close to the stage control of the whole piece, and can improve the positioning accuracy of the loading table. Here, the meaning of "weighting" is not only to select one of the slice-by-chip mode and the whole chip mode according to the linear error of the nonlinear error disk, but also to calculate the target position according to the magnitude of the nonlinear error and the linear error. . The method for exposing the shirt of the present month is to use a projection exposure device to irradiate the exposure light onto the patterned reticle to project the pattern image onto the exposure substrate loaded on the loading platform; the projection exposure device The utility model comprises: a driving stage, wherein the loading platform for carrying the substrate is moved and positioned in at least three predetermined positions; a position signal output mechanism is disposed on the driving stage, and outputs a position signal indicating a position of the loading platform; a detecting means for detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate; and a w position determining means for determining the loading stage based on the position indicated by the position signal 44 200907596 is characterized in that: the exposure substrate has at least three exposure regions for projecting the image of the pattern, and the substrate reference mark ' is used to indicate the exposure of the reference position of the at least three exposure regions. a region reference mark; and comprising the steps of: exposing a substrate position control step at which the exposure substrate is used as the base After being loaded on the loading platform, the loading stage is sequentially moved by the driving stage to be positioned at at least three predetermined exposure substrate detecting positions; and the exposure substrate position storing step is obtained from the position signals. Exposing the position of the loading table of the substrate detecting position and storing it; exposing the area reference mark position storing step m the exposure substrate=position '(4) the reference mark detecting mechanism detecting the exposure area reference 's' and then detecting the gentleman's fruit plate according to the exposure The position of the exposure &amp; field reference mark is calculated and stored from the position of the loading station, and the linear error correction operation step is performed according to the exposure area which is stored in the exposure area reference μ position storage step The minimum square method is used to calculate the position for the second position, and the material is used; the linear error correction component calculation step of only the stagnation tool is used to calculate the reference mark in the exposure region and the exposure domain reference The difference between the positions of the marks; the weight: control step, which is replaced by at least one of the least square red turn and the orthogonal The exposure = 45200907596 to Li exit ray 'Μ isobutyl controlling the pattern image and the projected image and only the exposure position overlaps the target substrate during the exposure of the substrate. Since the least square method _ &amp; π is used to take at least one of the expansion, rotation, and orthogonal of the 卞 凌 凌 置换 置换 , , , , , , , , , , , n n n n n n n n n n n n n n n n n n n The substrate can be positioned at a high speed at a high speed without using a position measuring device such as an interferometer for measuring the position of the stage. Further, even for a printed circuit board having a nonlinear error smaller than that of the wafer substrate, sufficient alignment accuracy can be obtained. Furthermore, even if a mark is provided outside the exposed area, the printed circuit board having a large nonlinear error can be aligned, and the processing performed by the substrate can reduce the wear of the mark outside the exposed area, and the plurality of overlapping exposures can be With the same-marking for alignment, the entire layer can be aligned. [Embodiment] Hereinafter, embodiments of the present invention will be described based on the drawings. <First Embodiment> FIG. 1 is a schematic view showing a projection exposure apparatus according to a first embodiment of the present invention. The projection exposure is i 100, mainly used to manufacture printed circuit boards. <<Configuration>> <Projection Optical System> <Light Source 11 0> The light source 110 is a light beam for emitting a desired wavelength. For example, it is possible to emit a short-wavelength ultraviolet ray for a mercury lamp or the like. In the bulb of the light source i 10, mercury as a luminescent material, and an anode (not shown) and a cathode (not shown) 46 200907596 are sealed. The above anode and cathode are arranged in opposite directions. Each electrode is electrically connected to a metal conductor (not shown) to form an arc discharge between the anode and the cathode. One of the jaws 112a of the light source 110 is fixed to a support member provided outside the ellipsoidal mirror 120 (not shown, and the other jaw 112b is connected to a power cord (not shown). The light source 11〇 The discharge is applied to the electrodes by passing a predetermined voltage through the jaws. Further, the small diameter of the power supply line does not interfere with the light beam emitted by the light source 11. After the arc discharge occurs between the anode and the cathode, A strong light is generated by a discharge portion called an arc column. The light beam emitted from the arc column becomes a divergent light that diffuses toward the square. <Elliptical mirror 120> The elliptical mirror 120' is used to concentrate the light beam emitted by the light source As shown in Fig. 1, the 'ellipsoid mirror 120 has a reflecting surface 122, and the elliptical mirror 120 has a shape of the reflecting surface 122 as a mirror for rotating the elliptical surface. A through hole 126 is formed at the bottom of the elliptical mirror ι2〇. A part of the light source 1 i is disposed in the through hole 126. By arranging a part of the light source 10 in the through hole 126, the light source 10 0 can be positioned such that the arc portion of the light source 位 is positioned in the elliptical mirror 120 The first focus. The fact that the light beam emitted by the light source 110 is temporarily concentrated by the elliptical mirror 12 可 can thereby increase the illuminance of the light beam. Thereby, the utilization efficiency of the light beam emitted by the light source 1 ^ 可 can be improved. The apparatus uses an elliptical mirror 120 in the concentrating optical system. The elliptical mirror i 20 has two 47th 200907596 focal points of a first focus and a second focus; the light source 110' is such that the electric part of the light source 11 位 is located at the elliptical mirror 120. The focus is supported by a support member (not shown), whereby the light beam emitted by the light source 110 is reflected by the reflection surface 122 of the elliptical mirror 12A and then concentrated at the second focus. 0> The optical rod 130 has an elongated rectangular shape. The optical rod 13 has an incident surface 132 and an emitting surface 134. The optical rod 13 is arranged such that the light source of the optical rod 130 is optically like a common surface. The second focus of the elliptical mirror 12 、 or the point of sharing with the second focus, and the arrangement of the optical rod j 3 〇 also causes the emitting surface 134 of the optical rod 13 , to be located later and later. The pattern forming surface ι44 of the reticle 142 is in a conjugated position .

On the incident surface 132 of the optical rod 130, the light beam emitted from the light source 11 is incident. The optical rod U is used to provide a nearly uniform illumination of the beam incident on the incident surface 132. From the exit surface 134 of the optical rod 13 ,, a light beam having an illuminance close to a uniform light can be emitted. The use of the optical #13〇 can be performed as long as the light beam incident on the incident surface 132 can be illuminated with a uniform illumination. <Light Guide Optical System 140 &gt; The light guiding optical system is composed of a mirror 136, an illumination relay optical system 138, and a mirror 140. The light beam emitted from the exit surface 134 of the optical rod 13 改变 is changed by the mirror 丨 3 6 _ brother U 6 , and the direction of the contract is enlarged by the illumination relay optical system 138, and the size of the c ^ stamp surface is enlarged. The direction of travel is again changed by the mirror 140, and is incident on the reticle 142 which will be described later. <Marking sheet 142> The marking sheet 1 42 is a type of 并, a, a gentleman &amp; cover for manufacturing a printed circuit board, 48 200907596 A conductor pattern is formed on an exposure substrate 156 (156a, 156b) to be described later. , or 156c). The reticle 142 corresponds to a negative film for transferring the pattern formed on the reticle 142 to the exposure substrate 156 by the exposure light emitted from the light source 110 to form a conductor pattern on the exposure substrate 156. The pattern formed on the pattern forming surface 144 of the reticle 142 corresponds to the conductor pattern to be formed on the exposed substrate 156. The exposure substrate 156 is composed of a substrate such as a copper laminate substrate, and is a substrate before the formation of the conductor pattern, and refers to a substrate on which a photosensitive material such as a photoresist is applied. Further, the conductor pattern refers to a pattern formed by a conductive material on a printed circuit board. Further, the printed circuit board means a conductor pattern having been formed on the exposure substrate 156. As described above, the light beam emitted from the emitting surface 134 of the optical rod 130 is changed by the reflecting mirror I4G, and is illuminated by the pattern forming surface 14 of the reticle (4). <The reticle shutter 146> A reticle shutter 146 is disposed near the upper side of the reticle M2. The reticle shutter 146 has a moving plate 148 which is movable while being parallel to the reticle M2. Further, in the example shown in Fig. 1, the reticle = the moving plate 148 which is held horizontally as the reticle shutter 146 is moved in the horizontal direction as indicated. The reticle covers... can be optical rod 13. A part of the light beam emitted from the exit surface 134 is of a desired size. The light beam after the cross section is changed, for example, is incident on the reticle 1 42 described above. <Projection Lens 150> 49 200907596 A projection lens 15A is provided below the above-described reticle 142. The projection lens 150 has an incident surface 152 and an emitting surface 154 which are supported by a supporting member (not shown) such that the incident surface 152 is positioned on the upper side and the emitting surface 154 is positioned on the lower side. The incident surface 152 of the projection lens 150 is incident on the light beam transmitted through the reticle 142. The projection lens 150 is composed of at least one or more kinds of lenses. In the inside of the projection lens 150, the cross-sectional size of the light beam incident on the incident surface 152 is converted into a desired size by the lens thereof, and then The converted beam is emitted from the exit surface 154. Thereby, the pattern formed on the pattern forming surface 144 of the reticle 142 to be described later is changed in size by the projection lens 150, and the pattern image is projected onto the exposure substrate 156, which will be described later, so as to be formed on the reticle. The pattern of sheet 142 can be transferred to exposed base 156 at a desired size. Further, the projection lens 15 () is preferably constituted by an equal magnification projection optical system. <Alignment Optical System 160> The alignment optical system 丨60 photographs a substrate reference mark formed on an exposure substrate 156 (156a, 156b, or 156c) or a reference substrate 158 (1583 or i58b) to be described later. The position for determining the substrate reference mark. The alignment optical system 16A is mounted on a moving stage (not shown) and is positioned at the retreat position and the measurement position. The retreat position and the measurement position will be described later. Further, the timing of the operation of the alignment optical system 丨60 is performed, and the table reference mark 179 of the turret (taMe) m formed on the substrate stage 17 (described later) is photographed to determine the table reference. When marking the location. 50 200907596 The μ alignment optical system 16G includes two microscopes 162a and 162b. The U-mirrors 162a and 162b each include a photographic element such as a ccd lens (not shown for photographing the substrate reference mark or the table reference mark. The photographic elements of the micro, brothers 162a and l62b are electrically connected to the image processing device (6). The image processing device 164 reads the image of the microscope 162 &amp; and (10) :: as image data, and applies image processing to the image data to extract the image of the substrate reference mark or the image of the table reference mark. To determine the position of the substrate reference ip &amp;# $ I . + 铋. The position of the own or the table reference mark and to save. 〃 者 ′′ In the example of Figure 1, there are only one alignment optical system 160, It is arranged in the projection lens! 5〇's station and #u Below the left side 'but' can also be equipped with 2 pre-systems to become the alignment optical system and secret. In the alignment optical system 16〇a盥The lower left system 160 of the month opening/projection lens is disposed in the same manner on the left side of the side, and the alignment optical system 160b is disposed on the lower right side (not shown) of the projection lens 15A. The alignment, the alignment optical system has Same composition Each of the two microscopes can have a shadow on the base reference mark of the substrate base mark. Further, the alignment optical system 16〇3 and the σ entrance pupil/a and 160b' are electrically connected to the image processing apparatus. It can be processed by the aligning optics "(10) &amp; and ... self-photographing. After using the above two alignments, the processing can be made more rapid. When the moon is shaped, such as <substrate loading station 170> As shown in Fig., the projection traverse 15 170. The configuration of the substrate loading table 170 is further loaded with the table 3, the X-direction moving cutting table 51 200907596, the direction moving stage 174, and the direction moving stage 176. In the figure of Figure 1, the direction of the inside of the circle of the six &amp; the direction of the circle to the right indicates the direction of +Χ, the direction of the ice to the surface of the figure is +Υ direction, Figure 7 is called the top of 07 In the direction of the 载-moving stage 176, the upper surface of the shape 水平1 is a table 177 (not shown) for loading the exposed substrate! Work:: It is in the upper part of the work ° 177, forming a working σ reference mark not shown] 79, used only m, 隹 (10) 俾 is used to form a reference mark formed on the reticle M2 for indicating the reticle reference mark, ...p position of the bar piece 42. When the reticle enters the 、~, machi image (below The position of the image called the base of the squad, and the position of the workbench + rank "W9" is aligned with the syllabus of the syllabus (10). The position of the storage table reference mark $HP is projected onto the exposure substrate by the pattern of the reticle 142. The use of the table reference for storing the first is defeated by the location of each other; the position of the position; The substrate reference pattern is formed on the desired exposure area of the reticle 142 to the exposure substrate 156. 'and the x-direction moving stage 172 and the Y-direction moving stage 174> Fig. 2(a) is a detailed front view of the X-direction moving stage 172 盥 γ square 74; Figure 2 (8) U-direction movement _ / 17-inch 2 ping front view. Further, in Figs. 2 (4) and (b), the figure shows the +X direction, and the upward direction of the figure shows the outer direction. The direction substrate loading table 170 has an elongated fixed plate ma and a fixed base portion (not shown). °疋盘 52 200907596 In the fixed plate 1 78a, with a linear motor stator! 8〇a and linear motor stator 1 82a. The linear motor stator 18 8a and the linear motor stator bore 82a are elongated. The linear motor stator 180a and the linear motor stator i 82a are disposed in parallel with each other and in the longitudinal direction of the fixed plate 178a, and are disposed on the fixed plate 178a. The fixed plate 1 78b is also provided with a linear motor stator 丨 8 〇 b and a linear motor stator 182 b. The linear motor stator 180b and the linear motor stator 18 are elongated. The linear motor stator 180b and the linear motor stator 1821) are disposed in the fixed plate 178b so as to be parallel to each other and to the longitudinal direction of the fixed plate 178b. Further, between the linear motor stator 18〇a and the linear motor die 182a in the fixed plate 178a, an elongated linear slider 188a is provided. Between the linear motor stator 180b and the linear motor stator 182b in the fixed plate 178b, an elongated linear slide 丨 88b is also provided. On the left side of the lower surface of the carriage table 74 for moving in the Y direction, a linear motor mover ma corresponding to the linear motor stator 18A is provided. Further, a linear motor mover 184b corresponding to the linear motor stator (10)b is provided on the right side of the lower surface of the γ-direction moving stage 174. The f-stream is supplied to the linear motor mover (4) or 184b, and the guide edge of the linear motor mover 184a is linearly slid along the linear motor stator 丨8Qa, and the linear motor mover 184b side is linearized. The leading edge of the slipper mb moves along the linear stator 1 8 0 b, and as a result of f φ shifting, the Y-direction moving stage 174 can be moved to the desired position in the ±Y direction. On the left side of the lower surface of the turbulent movement stage 174, a linear motor mover i86a is also provided corresponding to the linear 53 200907596 motor stator 182a. A signal indicating the position of the linear motor mover phase detector for the linear motor stator i 82a is emitted from the linear motor mover 186a. Further, on the right side of the lower surface of the γ-direction moving stage 174, a linear motor mover leg is provided corresponding to the linear motor stator 182b. From the linear motor mover, a signal indicating the position of the linear motor mover relative to the position of the linear motor stator is issued. * The γ-direction moving stage 174 is connected to the control device 199. The control ::199 can receive the signal from the linear motor mover and the signal from the linear horse 186b to obtain the gamma direction position or the moving distance of the Y-direction moving stage 174. When the ¥ direction movement stage 174 is moved in the ±Y direction, the control unit 199 receives the signal from the receiving line, and (4) issued by the Philippine V 5a i secret, according to the linear motor mover 186a and 186K. The position of +, to control the current supplied to the linear motor mover 184a or 184b. Further, when the linear motor mover 18 is moved in the Y direction, the movement in the Y direction is moved in the direction of (4), and even (4) is moved in parallel in the ±¥ direction. Further, the u movers 186a and 186b move at different distances, or the :Π:: square: the η: fly reverse clock direction has a certain degree of rotation. Fig. 2(b) shows the X-direction moving stage i 72 as shown in Fig. 2(b): #蒈古妗East The upper part of the Ganwan mobile station 174, the motor stator 190 and the linear motor stator 192. Soul | stator 19 〇 and linear motor; 1 192 192 ° linear motor sub-192 is elongated. The linear motor stator 19〇54 200907596 and the linear motor stator 1 9 9 # 2 are parallel to each other and move the longitudinal direction of the stage 174 in the γ direction from the initial use to the extension, and are placed in the γ direction. Taiwan] 74. In the moving square 1 74, an elongated linear slide 1 94 is provided between the linear motor fixed 1 90 and the linear motor stator 1 92. The lower side of the stage 172 is moved in the X direction, and the linear motor mover 196 is slid in a manner similar to the linear motor stator 19 &. The current is supplied to the linear motor mover 196' so that the linear motor mover 196 is guided by the linear slide 194, and moves along the line every ice, and as a result, the X direction can be moved. The stage 172 is moved to the desired position in the ±\ direction. A linear motor mover 198 is also provided on the lower surface of the X-direction moving stage 172 so as to correspond to the linear motor stator 192. From the linear motor: sub-198, a signal is generated indicating the position of the linear motor motion + 198 relative to the linear motor stator 192. The X-direction moving stage 172 is also connected to the control device 199. The control unit 199 can obtain the position or the moving distance of the direction shifting stage 172 by receiving the signal from the linear motor motion + 198. Further, as described above, since the Y-direction moving stage 174 can be rotated to some extent in the clockwise or counterclockwise direction, the signal 'supplied by the linear motor mover 198 indicates that it moves in the γ direction. The position of the stage m is moved in the X direction of the longitudinal direction of the carrier 174. When the 乂-direction moving stage 172 moves in the longitudinal direction of the γ-direction moving stage 174, the control device = 99 receives the signal sent from the linear motor material 198, according to the position of the linear motor mover 198. To control the current supplied to the linear motor mover 196. 55 200907596 The linear motor is activated by the signal from the gamma mover 186a and 186b as the home position of the 174 moving position. The position of the signal sent by the opposite party in the coordinate system. In addition, the start of the X-direction moving stage 1 72 by the linear motor mover-origin is used as the origin, and the coordinate system of the 乃^-Yimen mobile stage 174 which is fixed to γ as the origin. The position of the direction shifting stage 172 is 174 ^ e _ and the position is moved, and the position in the Y direction is: the position in the longitudinal direction. In view of the above, the "control device 19: = the upper part of the mobile stage 174" is issued by the '1', and the linear motor mover 1 86a and 1 86b are scared &amp; The position of the coordinate system of the | &amp; 178b can be converted to the position of the fixed load port 172. The control device 19 9 can receive the signal from the @ signal, the linear motor mover 186tH sub-speaker (9), and the linear motor movement "acquired in the fixed plate 17~ Or in the coordinate system of (10), v ^ a must be 4 ^ 1/50. 1 Use the stage 174 _ ¥ direction position or rotation position to obtain the coordinates fixed in the fixed plate 丨m turn ^ μ ί ι m ^ ^ or 178b In the system, the position of the gantry door 172 is moved in the direction of the direction of the stage 172. Fig. 3 is a side view of the Z-direction moving stage Α ^ , , θ coin battle α 176. Furthermore, in Fig. 3, the leftward direction of the drawing, γ^, 俜 Χ Χ Χ Χ a a ' ' 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深 深The position of the end position of the figure σ 4 is shifted in the direction of the position of the carrier 174 in the +Υ direction. σ system 56 200907596 The upper part of the stage 172 for moving in the X direction is provided with z-direction movement. The stage 176 is used. As described above, in the upper surface of the z-direction moving stage 176, a table-side moving load for breaking the exposure substrate 156 is formed. The micro-tilt drive motor (not shown) that can tilt the z-direction moving stage 176 horizontally is also possible to move in parallel in the ±z direction. The two directions can be controlled by the control of the micro-tilt drive motor. The table 177 of the moving stage 176 is inclined at a desired angle to the horizontal plane. As shown in Fig. 3 (4) and Fig. 3 (b), the working port 177 of the 纟Z direction moving stage 176 is loaded with the exposure substrate 156. The z-direction moving stage 176 is tilted by the above-described micro-tilt mechanism, so that the exposure substrate i56 of the table m mounted on the z-direction moving load 76 can be inclined at a desired angle to the horizontal plane. As shown in the items 3 (4) and _ 3 (8), the white arrow in the downward direction indicates that the beam is emitted by the emitting surface 154 of the projection lens 150 (not shown). The light beam emitted from the emitting surface 154 of the projection lens 150 is The "shot" exposes the substrate 156. The range of the arrows shown in Fig. 3 (4) and Fig. 3 (b) is the exposure area A which is exposed by the light beam emitted from the light source 110. It is preferable for the worker to use the flat table 177' of the stage 176 for the second t-moving stage 176. However, it is subject to the size of the work-light substrate 156, and the work has to be increased. The exposure of Taiwan 17^77. In this case, even if the work A 177 is in the shape of a work, there will still be a workbench; and 1J is well flattened. When the occurrence of the click, that is, the so-called deflection, as shown in the case where the table 177 of the Z-direction moving stage 176 is bent 57 200907596, when the exposure substrate 156 is loaded on the table I?? The substrate 156 is bent along the already curved table 177. In the case where the exposure substrate 150 is mounted on the upper blade σ丄77 in a curved state, as shown in FIGS. 3(a) and 3(b), it is possible to control the above-described ζ direction σ 76. The stage is tilted to drive the motor so that the table 丨 is inclined to a plane so that the exposure area eight in the exposure substrate 156 is horizontal. In addition, the control of the micro-tilt drive motor can also be performed by any of the closed-loop control circuit controls. Closed-circuit control means that /Kan AL does not move to the station = micro-tilt measurement sensor (not shown), and the micro-tilt drive is controlled according to the micro-tilt amount measured by the sensor in the micro-tilt. . On the other hand, the open circuit control means that the degree of the (fourth) curve generated by the table 177 of the Z-direction moving stage 176 is measured in advance, and the result is stored. When the exposure is performed, the pre-stored bending-end control micro-tilt drive is read. The motor is used to determine the inclination of the work 纟m. - The operation of mounting the exposure substrate 156 on the upper surface of the stage 176 for moving in the z direction as described above.

The opening 177 and the Z-direction moving load A / are provided on the upper portion of the X-direction moving stage π. Therefore, both of them are driven by the above-described X-direction movement stage 172# γ direction movement cutting a, and the operation 177 is positioned at the desired X-direction position: γ-direction position. <Abbe occurrence of Abbe error>: Although the tilting of the two-direction moving stage Μ can be performed, the exposure area A to be exposed is kept horizontal, however, the z-moving load 176 is tilted. In the case, the distance between the two-direction moving load : μ : the quasi-position and the surface of the exposure substrate 156 (two shown in FIG. 3 ) is changed by the inclination of the movement direction of the movement. Therefore, there is a case where the Abbe error occurs with the distance Δ. In the case of the Abbe error, the position in the horizontal direction is measured, and the size of the exposure substrate 156 mounted on the table 177 is different. There is a case where the moving distance of the X-direction moving stage 172 or the moving stage m has to be lengthened. Therefore, it is necessary to lengthen the linear slide 194 of the X-direction moving stage 172 and the "γ-direction movement". The horizontal length of the linear slides 188 &amp; 188b of the stage m causes the linear slide 194, or the linear slide 188a, im easy to have the z-direction bundle 1 linear slide 194, or the linear slide 188a, (10) The occurrence of the z-direction bending 'also causes the Abbe error in the z-direction to cause a measurement error in the horizontal direction. <Exposure substrate 1 56a and 156b &gt;&gt; FIG. 4 is an example of exposure substrates 156a and 156b Fig. 4(4) and the rectangle from the outside of the example A', showing the line segment of the outline of the exposed substrate core or 156b. Further, on the inner side of the outer rectangle, four vertical and three vertical directions are shown. square Eri to eri2), each of which indicates an exposure area, and transfers the pattern formed on the reticle to the 12 exposure areas ER1 to ER12. Further, it is used to indicate 12 exposure areas ER1~ The square segments of the respective contours of the ER 12 are used to show the hypothetical line segments of the exposed areas. When the exposure substrate 156a or 156b is loaded on the table 177, 59 200907596 is shown in Figures 4(a) and 4(b). Generally, the longitudinal direction of the exposure substrate 156a or 156b is the X direction of the substrate cracking stage 170; the lateral direction of the exposure substrate 156a or 156b is the Y direction of the substrate loading stage 170. Hereinafter, there is no special distinction between the exposure substrate 156a and the exposure substrate 156b. If necessary, it is referred to only as the exposure substrate 156. As shown in Fig. 4(a), the exposure substrate 1 56a in which two reference marks are formed in the exposure region is taken as an example; as shown in Fig. 4(b) For example, the exposure substrate 1 56b in which four reference marks are formed in the exposure region is taken as an example. As shown in Fig. 4(a), two reference marks RM1 are formed in each of the twelve exposure regions (ER1 to ER12). And the exposure substrate 156a of RM2. In the example shown in Fig. 4(a), in the exposure The two reference marks RM1 and RM2 formed in the respective fields ER1 to ER12 are located in the vicinity of the respective midpoints of the sides opposite to each other among the four sides of the square exposure region. Further, the formation of the two reference marks RM1 and RM2 The position is not limited thereto, and the two reference marks RM1 and RM2 can be formed by isolating the two reference marks RM1 and RM2 as much as possible. If the two reference marks RM1 and RM2 are formed as much as possible, the two references can be improved. Mark the position measurement accuracy of RM1 and _. In the example shown in FIG. 4(a), there are 1 exposure areas (ER1 to ER12) in the i-sheet exposure substrate 15, each of which is formed with two reference marks and RM2' in the i-piece exposure substrate 156a. Contains 24 fiducial markers. When the pattern formed on the reticle 47 &gt; Shishen 142 is transferred to the exposure substrate 156a, the reticle month used and the reticle reference mark are formed in the 〇 〇 月 month It corresponds to each of the two reference marks RM1 and RM2. As shown in the above-mentioned exposure substrate 156b of Fig. 4(b), four reference marks rmI to RM4 are formed in each of the 12 exposure 60 200907596 regions (ER1 to ER12). In the example shown in Fig. 4 (b), the four reference marks RM1 to RM4 respectively formed in the exposure regions eri to ER12 are located near the 4 正方形 of the square exposure region. Further, the formation positions of the four reference marks RM1 to RM4 are also not limited thereto, and can be formed by isolating the four reference marks RM1 to RM4 as much as possible. If the arrangement is as isolated as possible, the position measurement accuracy of the four reference marks RM1 to RM4 can be improved. In the example shown in FIG. 4(b), there are 12 exposure regions (ER1 to ER12) in one exposure substrate 156b, each of which is formed with four reference marks RM1 to RM4, and is contained in one exposure substrate i56b. 48 benchmark marks. When the pattern formed on the reticle 142 is transferred to the exposure substrate 156b, four reticle reference marks are formed in the reticle used, and each of the four reference marks RM1 to RM4 is associated with each other. <Displacement of Reference Mark> The reference mark RM1 formed on the exposure substrate 156a or the reference marks RM1 to RM4 formed on the exposure substrate 156b should be placed at a desired position when formed. However, the reference mark is The formation method γ is formed by exposing the surface of the substrate 156 by laser light, or is formed by machining such as a drill. Due to the formation method of the fiducial mark, the formation position of the fiducial mark occurs from the position at the time of design to the position after the displacement. Further, it is also possible to apply heat to the exposure substrate 156 or to transfer: the force is applied to the exposure substrate 156, and the entire substrate 15 is exposed. The file is deformed. In this case, the position of the fiducial mark will be 61 200907596, which is different from the originally designed design position.

Further, as described above, there is a case where the Abbe error is caused by the X γ-direction moving stage 174 or the two-direction moving operation, the stage 172 or . The configuration of the carrier 176 is as described above. When the reference mark is set from the original position, $ θθ, and the position at the time of the rush, the displacement or the occurrence of the Abbe error occurs, the image is formed on the reticle 142. It is true that only the transfer to the exposure substrate is used to cut a, and it is necessary to move the load D 172 in the positioning X direction, and to move the 176 in the Y direction to the 176+卩*丄&amp; 4 or the Ζ direction movement port. In the case of taking the reference mark, the positioning is to be described later. Implement positioning. In the case where the reference mark is originally designed by the original position: to the position after the displacement, the position of the reference mark also occurs. <<<Transfer order of the pattern formed on the reticle 142> As described above, the pattern formed on the reticle M2 is transferred to the exposure substrate 156 by the exposure light emitted from the light source m. In the present embodiment W, the earth-boring plate 1 56 has a plurality of exposure regions, and the pattern formed on the reticle 142 is sequentially transferred to each of the plurality of exposure regions. Before the pattern formed on the reticle 142 is transferred, the above-described pre-preparation *Μ乍 must be performed so that the position of the image of the I-line reference mark and the position of the table reference mark 179 coincide with each other. On the upper side of the table m, a table reference mark 179 is formed for overlapping the position of the reticle reference marks formed on the reticle Μ]. The yaw direction movement stage 172 and the Υ direction movement stage m are driven to move the operation from 177, and 62 200907596 causes the position of the reticle reference mark to be accurate. When the Iraqi line is only set to 'can be marked with the workbench reference mark 179 mark 丨79: the position of the image of the two light reference mark, the workbench reference mark of the workbench base, the alignment light (four), the system (10) The position of the mark 179 is marked. The heart photography 'before the storage table base is to be formed when the reticle is first moved to the X-side pattern to be transferred to the exposure substrate 156 - to make the desired: two:: stage 17... The work stored in the preparation stage: two The position of the fiducial mark is the same as above. The positions of the σ-reference marks 179 are, for example, used to move the X-direction when the pattern is transferred to the exposure area 156m s indicated by the reticle 142 and transferred to the 177-th light area ER1 used in Fig. 4 (4). The stage 172 and the gamma directional direction of the 矽 矽 光 区域 区域 区域 h 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 Furthermore, as described above: when the 2 system is transferred to the exposure substrate, the reticle Φ used is a reticle having two reticle reference marks, and is specifically adapted to correspond to the two reference marks. Delete with _. In the same manner, when the exposure MER6 of the stage for moving the X-direction moving table shown in FIG. 4(b) is formed on the reticle 142, the direction/movement stage 174 is moved to ^ 'The position of the reference marks RM1 to RM4 of the exposure area ER6: the positions of the table reference marks 179 stored at the time of the process are the same as the above." In the case of transferring the pattern to the exposure substrate 63 200907596 (10), The reticle 142 used is a reticle that forms a fiducial mark, and is made to correspond to = RM1 to RM4, respectively. w. The moving X-direction moving stage 172 and the ¥ side 174 are displayed in the morning, and the desired exposure door moving stage is used as the reference mark 149 at the position of the reference mark and the stored work is exposed. To set = by 'light source 110 substrate 156. The pattern formed on the core piece 142 is transferred to the position where the exposure of the reference mark i of the position of the reference mark of the desired exposure area is not stored, and the exposure light is staggered, and then the light source U 〇 The pattern of the hair piece (4) is sequentially rotated, and this step 'can be formed on the reticle 、', and the sequence is transferred to each of the plurality of exposure areas. On the other hand, an error occurs in the position of the reference mark on the right side, and in the shape, the stage 172 and the shift = 174 must be positioned corresponding to the generated error. Directional movement of the stage < &lt;&lt;&lt;&lt; Alignment >> > The direction in which the X-direction movement is to be positioned in accordance with the error: :... The direction of movement of the stage 174 is (4). Piece-by-slice=die-die mode and whole-piece mode, for multiple exposure zones

Fang: (4) The position error, and then correct the X universal moving load a β Λ according to the error, thereby forming the positioning position of the moving carrier 174, on the one hand, the pattern transfer of the dead line 142 To the exposure substrate 156. Another error. ^ ^ means that the overall error of the exposed substrate 156 is pre-calculated across the entire substrate 156, and the error is measured for the position of the X side based on the error of 64 200907596: moving the stage with the stage 172 and the Y direction 174 is formed on the reticle 142: 'and the corrected position is stored in advance. When the transfer of the shape-moving carrier is transferred to the exposure substrate 156, the position of the crucible is sequentially positioned, and the carrier is transferred to the exposure substrate 156 after being positioned and corrected. <Paging method> &gt; Each finger of the exposure area, a plurality of references for the exposure base &amp;156; ^ 己 罟 3 W The alignment optical system 1 60 measures the position Γ According to the measurement As a result, the pattern of the "base" and the reticle 142 is transferred to the exposure substrate 156. The position of the substrate loading table 170 is corrected according to the measurement result by the measurement result of the position of the one placed by the alignment optical system 16A. Since the positioning of the directional movement stage 172 and the γ-direction movement stage 174 is not used by the interferometer, the substrate loading stage 17 is implemented. The positioning makes it possible to make the projection exposure 100 cheaper. For the piece by piece method, the details will be described later. Furthermore, in the first aspect to the third aspect of the step by piece method described below, the above-described pre-preparation is performed, and the image position of the reticle reference mark and the position of the table reference mark 179 are made. The position of the table 179 is stored in advance. <The third aspect of the step by slice method> The first aspect is to transfer the pattern formed on the reticle U2 to the 12 exposure areas eri to eri2 of the exposure substrate 15 6 &amp; Every 65 200907596 one. As described above, two reference marks are provided in the two exposure areas ER1 to ER12' of the exposure substrate 4 and the enamel plate 156a. In this first aspect, the alignment optical system 16A is used to determine the positions of the two reference numerals s1 and RM2. As described above, the I β alignment optical system 16 includes two microscopes 1 62a and 1 62b. The reference mark RM1 of the exposure substrate 156a is detected by the display mirror 162a, and the reference mark rm2 of the exposure substrate 1 56a is detected by the use of the mechanical mirror 1 62b. The two microscopes (6), 62b each include a photographic lens (not shown) such as a CCD lens, and the surface of the exposure substrate 156a including the reference mark RM1 or the photographic element is photographed, and the reference is taken from the captured image. Mark the image of RM1 A RM2 'to determine the position of the benchmark mark coffee or RM2. Figs. 5(4) to 5(4)' show the positional detection of the reference lever in the first aspect, the positional adjustment of the exposure substrate 156a, and the order of exposure to the exposure substrate 156a. Further, in the examples shown in Figs. 5(4) to 5(4), eri is represented by eri in the 12 exposure regions ER1 to ERi2 of the exposure substrate 156a. Further, in the example of the items 5(a) to 5(c) + the X direction, the left side of the drawing indicates the +Y direction. Further, the large circle ' in the example of Fig. 5(a): = indicates the illuminable range EA of the light beam emitted by the projection 5G. The state shown in Fig. 5(a) is driven by the X-direction moving stage 172 and the γ-direction moving stage m of the substrate loading table 17A, and is exposed to the exposure of the table 177. The exposure area ERi of the substrate 156a is positioned in the illuminable range EA0f. Further, the two reference marks _ and _ formed in the exposure areas 66 200907596 ER1 to ER12 of the exposure substrate 156a are stored in advance, and the y-direction movement stage 172 can be moved based on the numerical value. The rubbing direction moving stage 174 is positioned to expose the exposure areas ERi to ER12 of the exposure substrate core to the illuminable range ea. In the state of Fig. 5 (4), the alignment optical system and the system i6 are in position. Moreover, the retreating position of the alignment optical system 16〇, as shown by the circle, positions the alignment optical system 16 to be retracted from below the projection lens 15〇 to avoid alignment of the optical substrate. The exposure of the heart is carried out or each of them is obstructed to the exposure operation. For example, the retreat position of the alignment light system 260 is located, when the light source emits exposure light, the pattern formed on the reticle 142 is transferred to the exposure substrate ―, the subsystem 160 is positioned. position. The position, preferably in the Alignment Optics Department, is retracted from::) for simplicity, and its state is illustrated as: Light: System 160 is positioned by exposure "Alignment Said, the position where the 4 ER1 retreats, however, the actual position as above refers to the position after the optical system 60 is retracted. ', the substrate loading stage 170 is aligned with the optical system 160, including the x direction. The moving carrier is aligned by the X direction. (The white arrow ΑΧ1 or ΑΧ2 is not shown... Figure 5(a) or Figure 5(c) moves in the ±X direction. No, the alignment optical system 160 can be shown in Figure 5(b). In the state of the quasi-position, as described above, the alignment is first:: 2 〇 moving positioning is used to measure the quasi-moving stage, and by the symmetry of the x-direction; &quot; The drive of the stage, 67 200907596 The white arrow Αχι+ as shown in Fig. 5(a) can move the alignment optical system (10) in the X direction and is positioned at the measurement position. The measurement position is aimed at the hurricane. For the A 俨 先 先 系统 系统 系统 〇 〇 〇 〇 〇 〇 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜 显微镜It is only necessary to use the position detecting device: the 疋 alignment optical system, the position of the system 16 is positioned at the measurement position, that is, the alignment optical system 16 is rotated from the above-mentioned position. When leaving and positioning to the measurement position of the == mark RM1 or RM2, the measurement position should be continued at a certain position. The alignment optical system 16 is moved from the retracted position to the measurement position by the following method. First, for example, when the retreat position of the alignment optical system 6 is aligned with the start origin of the movement stage, The control unit for the motor for (4) X-direction alignment shift (four) stage pre-stores the number of pulse signals, and the number of the pulse signals corresponds to the distance from the retreat position to the measurement position. Apricot and scorpion from the field to the ',,, first 160 left from the retreat position and positioned in the measurement position

«7 I % Daily Call, J The number of the pulse signals pre-stored, from the control mechanism, the pulse number 彳', to the motor, thereby aligning the alignment optical system 1 60 from the X direction to the mobile load After the start point of the station leaves, it is continuously positioned at the position of the measurement. * As shown in Fig. 5 (b), the alignment optical system 16 is moved and positioned, and the reference mark RM1 is photographed by the microscope 162a, and the reference mark RM2 is photographed by the microscope 162b. Store the captured image in the image 68 200907596 Storage mechanism (not shown), at the end... take out the base 22: f = structure (not shown) to image % ® as early as 圮 RM1 and RM2, | The position of the ί^ΜΙ and the position of the reference mark. Furthermore, at this time, the position of the mark RM1 and the reference mark RM2 are accurate., r J. The position on your shirt image, for example, 'is only positioned in units of pixels (pixels), etc. The length unit used is used for positioning. ..., the position of the position _ of the reference mark coffee obtained in the above manner is the position in the X direction (the yoke mark square 2 = the reference mark R... the coordinates and the sill bar, the two coordinates, such as As will be described later, information on the deviation of the X from the summer, the amount of deviation, the degree of rotation, and the magnification can be obtained. The information of the four items will be described later. Further, the reference information mark RM1 for the four items of information waiting for it is further described. The location and benchmark mark coffee work ^ quasi / off, to understand whether it is pre-stored in the pre-preparation phase of the benchmark private § 179 whether they are in position to each other. Save the workbench benchmark mark 179 ... pre-reserved performance The drawing is formed at the position of the image reference mark (10) of the image of the line reference mark of the reticle 142, and the position of (4) m^RM1 and the position i of the mark 179 are not made "pre-stored table reference.

For the direction movement: Then, the substrate loading table 17 is driven. X 々 移动 移动 移动 太 太 太 a a a 基板 基板 基板 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 、 、 、 、 、 It can be compared with the pre-existence, and it can be used to mark the position of the RM1 and the reference mark rm2 69 200907596 with the reference mark rm2 69 200907596. The reference is recorded in the position of the 铋 士 式 , , , 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 曝光 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 In this case, it is also possible to control the magnification change of the projection lens 150, and the magnification in the X direction is shown in Fig. 5(C), which means that you will be yelled again. 々 作 予 、 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , As shown in the mouth - Χ 2, it can be aligned # | will be transferred to the direction of _ 会 ( 四 四 四 四 四 四 四 旱 旱 旱 旱 旱 旱Further, the position of the silver and silver is the same as the above, 俜, so the stone is removed from the β to avoid the alignment optical system 1 60 becomes the exposure substrate 156a to expose the exposure to ^ ^ &amp; Shi Lijun et al. The obstacle is such that the alignment optical system (10) is located below the projection lens 15 。. The alignment optical system 160 is retracted from the lower side of the projection lens (9), tfc ' JS 1 1 Λ The exposure light is emitted, and the pattern formed in m is transferred to the exposure region ER1. Μ = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = The exposure area ER2 of the white exposure substrate 156a is positioned in a state shown in Fig. (4), and then transferred in the same manner as the above-mentioned step, the pattern transfer of the reticle c s 7 ^ which is formed on the reticle 142 To the exposure area ER2. By this, each pattern of the coffee ~ £1112 is transferred to 12 exposure areas (the second aspect of the order by slice method) 200907596 This second aspect will be formed on the marking The pattern of the sheet 42 is transferred to each of the twelve exposure regions ER1 to ER12 of the exposure substrate 156b. As described above, the four exposure regions ER1 to ER12 of the exposure substrate 156a are formed with four reference marks rmi to RM4. In the second aspect, the alignment optical system 160 is used to determine the four reference marks RM1. The alignment optical system 16A in the second aspect includes two microscopes 162a and 162b as in the ith aspect. The microscope 162a can detect the reference mark RM1 or RM3 of the exposure substrate 156b, and the microscope 162b can detect the reference mark RM2 or RM4 of the exposure substrate 156b. Due to the composition and function of the alignment optical system 160 and the first! Since the aspects are the same, the description is omitted in the second state 2. The two microscopes 162 &amp; 162 and 162b each include an imaging element (not shown) such as a CCD lens. The surface of the exposure substrate (4) including the reference mark 2 RMUM4 is imaged by the imaging element, and the reference mark is taken out from the captured image 7 The image of RM1~RM4 is used to determine the position of the base RM1~RM4. 6(4) to 6(4) show the position detection of the base in the second aspect, the position adjustment of the exposure substrate 156b, and the order of exposure to the two substrates (10). Ride, as shown in Figure 6 (4) ~ Figure 系), to expose the substrate (four) of the 12 exposure areas yang ~ coffee 2 as its representative. Further, in the examples of Figs. 6(4) to 6(4), the U direction of the drawing surface and the left side of the drawing surface indicate the +Y direction. Further, the large circle ' in the example of : indicates that the projection range 15 is emitted by the projection lens 15 (4). In the state shown in Fig. 6(a), the drive unit 172 and the Y-direction moving stage 174 of the substrate loading table 17 are driven to be mounted. The exposure area ER1 of the exposure substrate 156b of the stage 177 is positioned at the illuminable range EA. Further, in the same manner as in the third aspect, the four reference markers "丨~尺河彳" formed in the exposure regions ER1 to ER12 of the exposure substrate 156b are designed to be previously stored and can be moved according to their values. The X-direction moving stage 172 and the Y-direction moving stage 174 are positioned to expose the exposure areas ER1 to ER12 of the exposure substrate 156b to the illuminable range EA. In the state of FIG. 6(4), the alignment optical system 16 is ligated. In the retreat position, the retreat position is also the same as the first embodiment. The state shown in Fig. 6(b) is that the alignment optical system 16 is moved and positioned at the measurement position 'the substrate 丨56 is also moved. To the entry and exit position. The same as the first aspect, the 'alignment optical system 16' includes the χ mobile stage', as shown by the white arrow AX1 of the drive figure 6 (4) for aligning the Χ direction with the movement stage. The alignment optical system + direction is moved to be positioned at the measurement position. The position of the measurement bit 1 is different from the line position (refer to FIG. 5 (10), however, only the second and the younger can make the microscope 162a to the reference mark RM1 "criminal dimension can also make the microscope 162b against the benchmark It is sufficient to set the position of RM2 or 4, and to set the position. In the other aspect, the direction of the 蹑 蹑 蹑 , , , , θ θ θ 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 Shifting &quot;, 172 moves in the -X direction, and is switched from the original position to the station and the position. 72 200907596 Furthermore, the original position of the loading table 172 in the χ direction is called 6 (c Or the position shown in Fig. 6(d). From the original position _. The movement of the home position can be performed as follows. First, in the control mechanism of the motor for controlling the turret of the turret, Pre-stored the number of the pulse = shift number, and the number of the pulse signal corresponds to the distance from the original position of the moving direction &amp; - 2 172 up to the entry and exit position. Then: ::: direction moving stage The 172 is moved from the original position to the inbound and out, and the pulse signal is supplied from the control mechanism (to the motor according to the age of the previously stored pulse signal, thereby enabling the 乂-direction moving stage u to be rotated and positioned to the in-and-out position. The original fruit position moves the alignment optical system 160 to the measurement position, and In the X direction, the movement of the stage 172 is moved to the push-in &amp; the ^^ can be used to position the alignment optical system 160^° from the 172 as shown in Fig. 6(8), even if it is based on the whole: Z RM1 is the microscope 162a The photographing area includes: The whole of the soil mark RM2 is included in the photographed area of the microscope. As shown in (6b), the alignment of the buds to the drought system 160 is positioned at the measurement position to move the X direction. The stage 172 is positioned in and out of the reference mark RM1, and is imaged by the microscope 162b. The image to be captured is imaged by a video storage mechanism (not shown): = "bad" "image processing mechanism (not shown) In the image processing, the images of the reference numerals RM1 and RM2 are taken out to calculate the position of the bit mark RM2 of the reference mark leg i. Furthermore, 'the position of the reference mark of the base drought reference mark ^^2 and the pixel (pixel) etc. as the unit to determine the position of 'i, for example, as long as it can be used in the early position, it is not necessary Use millimeters, etc. - General Office 73 200907596 to use the length unit to locate. The alignment optical system 16 is maintained in the state of the month in which the X-direction movement is moved in the direction in which the vertical direction X is moved back to the 々 (4), and is positioned at the measurement position. By aligning the optical system (10), it is possible to use the pair of reference marks RM3 to photograph, the 镜m mirror 162b, and the reference mark "missing the image to the image storage machine". It is to store the camera (not shown 亍t: ( ), and then use the image processing _ Ϊ processing, take out the reference mark _ and the image to calculate the position. Go again, "The position and the reference mark are compiled this % The rice 1 of the reference mark RM3 is also after the hit... The π position is positioned in units of the reference mark RM4, and the stand-up 'for example, as long as the pixel (pixel) can be used as a ^, it is not necessary to use a millimeter or the like. The position of the four reference mark faces to 譲4 obtained by the length single description method, the position of the ° (x coordinate) and the position of the YS direction (γ coordinate). The position of the leg is marked by the material reference mark, (4) The X-displacement amount, the γ-deviation amount, the rotation degree, the magnification in the χ direction, and the magnification in the γ direction of the exposure region are obtained as information of five items. Further, the information about the information of the exposure region will be described later. Judging the position of the reference marks RM1 to RM4 to understand the pre-preparation stage The pre-stored work 179 θ 3 ^ The position of the pre-stored workbench reference mark is the reticle reference mark of the pre-stored reticle 142 The position of the image is 74 200907596. If the position of the reference mark is read, the position of the reference mark is judged;:::7: the position of the reference mark 179 is not - and it is the drive table of the X direction of the drive. The position of the exposure area ε of the exposure substrate 156b and the γ direction (4) 174 can be positioned with the pre-stored operation c. Thus, the bits of the reference marks RMI to RM4 are formed on the reticle. The position of the reference mark of 142 corresponds to the position of the exposure area ER1 of the plate 156b. Further, if it is the magnification in the X direction and the Y direction obtained from the position of the reference rank, the leather &quot; In this case, the control of the rate change can be simultaneously performed so that the magnification in the X direction and the γ direction are doubled. Fig. 6(4) shows the state in which the alignment optical system (10) is again moved to the V position. The above-mentioned χ direction is aligned with the moving load «' as shown in Figure 6 (4), the white arrow ΑΧ 2 In the same manner, the alignment light can be moved, and the system is moved in the direction of the retreat. In addition, the retreat is the same as described above in order to avoid making the alignment optical system 160 Obstacles to exposure or various operations, etc., so that the alignment photonic system 160 is located below the projection lens 15 。. ρ 光学 optical "(10) self-projection: after the lower retreat] is emitted by the light source 110 The pattern of the sheet 142 is transferred to the exposure region ER1 by using light. The line is followed by the exposure of the substrate loading stage to the Y-direction moving stage 174. 156b, the field ER2 'positioned in the illuminable van 75 200907596 II: is: two: shows the state, and then in the same step as above, can be transferred to the exposure area ER2 according to the pattern of the sequence 2. Thereby, the pattern of each of the sheets 142 of the coffee to coffee is transferred to the 12 exposure areas <the third aspect of the order in which the sheet method is performed> the third aspect 'transfers the pattern formed on the reticle 142 to Each of the twelve exposure regions ER1 to ER12 of the exposure substrate (10) is exposed. The twelve exposure areas ER1 to ER12 of the exposure substrate 156a are opened; and T are four reference marks RM1 to RM4. In this third aspect, the detection is made: quasi-optical system (10) and 16 〇 b. The reference mark RM1 is excavated by the microscope 162aa, and the reference mark rm3 is detected by the microscope 162ab to detect the reference mark rm3, and the reference mark RM4 is detected by the microscope. The alignment optical system (10) and the configuration and function of the secret are the same as the alignment optical system 160 of the first aspect or the second aspect. Therefore, the description of the third aspect will be omitted. Each of the four microscopes I62aa and i62ab and 162ba and 162bb includes a photographic element (not shown) such as a CCD lens, and the surface of the exposure substrate (10) including the reference mark RM is photographed by the photographic element, and the image is taken from the image. In the middle, the reference mark RM1 to the image of the object is taken out, thereby determining the positions of the reference marks RM1 to RM4. Fig. 7 (a)' (7) shows the procedure for detecting the position of the reference mark by the third aspect, the exposure of the substrate 156b, and the step of illuminating the substrate 156b. Further, in the examples shown in Figs. 7(4) to 7(c), eri 76 200907596 among the twelve exposure regions ER1 to ER12 of the exposure substrate 156b is represented. Further, in the examples of Figs. 7(a) to 7(), the upper side of the drawing indicates the +χ direction, and the left side of the drawing indicates the +γ square_ Π Π, in Fig. 7(a) to Fig. 7 ( c) The large circle in the example also indicates the illuminable range 光束 of the beam emitted by the projection through the lens. The state shown in Fig. 7(a) is driven by the X-direction moving stage 172 of the Ershitian substrate loading table 170 and the γ-direction moving tens of the second-stage moving stage 174. The exposure of the station 1 77 positions the exposure area ER1 of the earth plate 156b in the illuminable range EA. Furthermore, the four reference marks formed on the exposure areas EIU to ER12 of the exposure substrate 156b are respectively disposed, and the design is placed in advance. The X-direction moving stage can be moved according to the value. 1 72 and the Y-direction moving load a σ 74 ' are used to position the exposure regions ER1 to ER12 of the exposure substrate 156b to be illuminable. In the state of Fig. 7 (4), the alignment optical system is in the retracted position. This retreat is also the same as the first or second aspect. As shown in Fig. 7(b), both of the singular semi-lighting systems 16A and 160b are moved and positioned at the measurement position. The alignment optical systems 160a and 160b each include a X-direction advancement mobile stage. By aligning the driving of the τ+ moving stage with respect to the X direction of the alignment optical system 160a, as shown by the white arrow AX5 in the figure, the alignment optical system 16〇a can be moved in the _χ direction. Positioned at the measurement position ν, ν, and again, as shown by the white arrow ΑΧ6 in Fig. 7(a), the optical wire can be aligned, and the cymbal 160b is moved in the +X direction to be positioned at the measurement position. The positioning state of both the optical system 16〇aW is aligned by the method of moving the lifting optical system 16〇a and the secret side to the measurement position, such as 77 200907596. FIG. 7(b) shows the entirety of the reference mark RM1 by the microscope 162aa. The photographing area is included; the entirety of the fiducial mark RM2 is included in the photographing area of the microscope 1 62ab; the entirety of the fiducial mark RM3 is included in the photographing area of the microscope 1 62ba; and the entirety of the fiducial mark RM4 is included in the photographing area of the microscope 1 62bb . As shown in Fig. 7(b), the alignment optical system 16a is moved and positioned at the measurement position, and the reference mark rmi is photographed by the microscope 162aa, and the reference mark RM2 is photographed by the microscopic 'brother 1 62ab. Similarly, the alignment optical system 16〇b is positioned at the measurement position, and the reference mark belly 3' is photographed by the microscope 162ba, and the reference mark is photographed by the microscope 162bb. The captured image is stored in an image storage mechanism (not shown), and image processing is performed by an image processing mechanism (not shown) to extract the reference mark RM1 to RM4 of the reference mark weiwei 4 Tus RM...4 is the position of 4. In addition, at this time, the reference marks KM1 to RM4 are in the first place, and the other is "A", the position of the shirt, for example, as long as the pixel (like /... positioning, no need It is positioned in the long-term production unit generally used in millimeters, etc., and the position of the four reference marks RM1 to RM4 obtained by the method is the same as the position of the X direction (x coordinate) and the third aspect. Set (Y^). By the position of the I and 4 reference marks RM1 to RM4, the X deviation amount and the γ deviation amount of the exposure area can be set, and the magnification of the X direction and the month V+丄γ deviate from 1, red. The degree of rotation, the right M direction, and the information of 5 items. Go to the 5 items about it and let it be described later. The poorer ones will be in the position of the benchmark mark to understand its position with the preparation Μ4 To make a judgment, it is said that the pre-spring stage is pre-spring and the pre-stored workbench benchmark mark 78 200907596 ^ The work of the redundant hoof is a bit L of the Huai mark 179 formed on the reticle base of the reticle (4); t : like the position. If the position of the reference mark coffee ~ circle ^ and the position of the pre-stored workbench reference mark 179 does not drive the base The X-direction moving stage 172 of the loading table 170 and the ¥-moving stage 174 are used to position the umbrella basin. The exposure area of the erbium substrate 156b is exposed to the position of the front surface of the substrate 156b. The pre-stored table bases 179 are positioned at the same position. Thereby, the reference marks RM1 to RM4 can be set: the reference marks formed on the reticle 142 are mutually aligned, and the positioning is based on the reference. ...向::: The position of the second 4...^ The case of the projection lens 15〇^^ 1 can also be used to control the magnification of the Y direction at the same time, so that the magnification in the x direction and Fig. 7(4) are * Once again, the state of the light m and the movement position at the retracted position will be aligned. By aligning the X direction of the above _ with the drive of the moving stage, as shown by 糸, = head AX7, Positioning the optical system μ ^ ^ arrow in the retracted position. Similarly, by moving the above:: toward the movement, the direction of the 160b is aligned with "the quasi-optical system head is like the movement" as shown in Figure 7. (c) The white arrow is positioned to retreat from the 1st and then the light (4) to move in the direction of the 16Gb ,, so as to avoid the alignment of the light wind season. Similarly to the above, the exposure optical system (10) 3 and 79 200907596 16〇b are located below the projection lens 150 in order to perform exposure or various obstacles such as exposure to the substrate. After the alignment optical system 16 is retracted from the lower side of the projection lens 15 as described above, the exposure light is emitted from the light source 110, and the pattern formed on the reticle 142 is transferred to the exposure area ER1. D, 17 ^ In the X-direction moving stage Δ"Y-direction moving stage 174 of the substrate loading table 170, the exposure area ER2 of the exposed exposure substrate 156b is positioned in the illuminable range A. The state shown in (4) is shown. Then, in the same manner as the above-described steps, the pattern of the slab 142 is transferred to the exposure region ER2. Thereby, the pattern formed on the reticle 142 can be sequentially transferred to each of the domains ER1 to ER12. In the third aspect of the nine-field, for each exposure area can be divided into = reference mark - "position, so that the detection can be two: 1: 1 and the third aspect, for each exposure area can take X The magnification error of the direction and the magnification error of the 丫 direction are both significant. Therefore, there is a significant effect on the magnification error of the individual γ-weighted x-direction and the magnification error of the Y-direction. < <Whole film mode> As described above, The whole piece method refers to, and the error is calculated, and the load is moved according to the measured overall direction of 56: the positioning position of the two-way moving stage Jiang Yang A dynamic load port 174 is pre-applied, and the subsequent position is given a mistake. Therefore, when the pattern formed on the reticle 142 is transferred to the exposure substrate 156, the #心线片 142 calls for positioning the X-direction moving stage 1 72 200907596 and the γ-direction moving stage 174 at After the correction, the pattern is transferred to the exposure substrate 1 5 6 in sequence. <Exposure substrate 15 6c> Fig. 8 is an example of the substrate (5) which is exposed in the case of the whole film mode. The exposure base shown in Fig. 156c, the above exposed substrate core and 15 6b is the same, the outer rectangular shape indicates the line segment of the outer shape of the exposed substrate 15&amp; and, on the inner side of the outer rectangle, four horizontally three vertical squares (ER1 to ER12) are illustrated, which are respectively indicated. One exposure area is obtained by transferring the pattern formed on the reticle 142 to the 12 exposure areas ER1 ER ER12. Further, a square line segment for indicating the respective contours of the 12 exposure areas ERhERK is used to show When the exposure substrate 156c is loaded on the table ι77, as shown in FIG. 8, the longitudinal direction of the exposure substrate 156c becomes the χ direction of the substrate slab 17 ;; the lateral direction of the exposure substrate 156c, The gamma direction of the substrate loading table 17 is formed. In the exposure substrate 156c shown in Fig. 8, two reference marks RM1 and rM2 are formed in each of the four exposure areas ER1, ER4, ER9, and ER12. The four exposure areas ER1. ER4, ER9, and ER12 are the exposure areas of the four sinus exposure substrates 156c. When using the exposure substrate 1 56c shown in Fig. 8, the two alignment optical systems 丨6〇 are used to determine two. Benchmark mark R The position of M1 and RM2. The alignment optical system 160 includes two microscopes 162a and 162b. The reference mark RM1' of the exposure substrate 156c is detected by the microscope i62a. The reference mark function of the exposure base 81 200907596 board 156c is detected by the microscope 1 62b. The configuration of the alignment optical system 16G is the same as that of the optical system 16G. The first aspect of the sequence of X卩 is the same 'thus, the reticle 142 used when it is to be formed on the reticle 142形成 = exposure substrate adder, 俾 and 2 reference marks and positions: position of the T-plate reference mark - 致 ==:: image position and ... reference mark _ The position of the reference mark (7) is stored in advance. The design position of the reference mark surface formed by each of the four exposure areas ER1, ER4, ER4, and ER12 of the exposure substrate 156c is 0 mark_deSl(n), Y mark-(4)(4)), and the design position of the reference mark 2 (X) Mark_deS2(n) and Y mark_des2(n)) are stored in advance, and the X-direction moving stage m γ-direction moving stage can be moved according to the value to expose the exposure substrate 156 to the four exposure areas. , coffee, and ER12^ to the illuminable range EA. Further, the variable n is to be described in detail as 0. <Position detection of the reference marks RM1 and RM2> Fig. 9 is a flowchart showing the processing of the entire substrate using the exposure substrate 15 &amp; The following description will be made using the flowchart of Fig. 9 . Further, the variable η' shown in the flowchart of Fig. 9 indicates the above-described four exposure regions EIU, ER4, ER9, and ER12. When n = l ' indicates that the exposure area is ER1, the exposure area ERi is referred to as the measurement exposure area of the first item. When n=2, when the exposure area is ER4, the exposure area ER4 is referred to as the measurement exposure area of the second item. n=3, is a table 82 200907596 = When the exposure area is ER12, the exposure area X is called the third item, the exposure area. When n=4, when the exposure area is ER9, the exposure area ER9 is referred to as the measurement exposure area of the fourth item. 1. First, the alignment optical system 丨6〇 is moved away from the retracted position and then moved to the position of the measurement (step S11). The retreat position or the ? 疋 position of the alignment optical system 160 is at the same position as the slice-by-chip mode. The preferred position of the retreat position is in the alignment origin of the alignment optical system. X, the setting of the position f is set as long as it can ensure that the position can be photographed by the microscope 162a. It is sufficient to position one of the photographing reference marks RM2 by the microscope 162b. Then, the X-direction moving stage 172 and the γ-direction moving stage 174 of the substrate loading table 170 are driven to move the stage 177 so that the exposure area of the exposure substrate 156c mounted on the stage 177 is positioned in the illuminable range. EA, 俾 is the measurement exposure area (4) of the first item (step S12). As described above, among the four exposure areas EIU, ER4, ER9, and ER12, the design positions of the reference mark RM1 (Xmark_desl(n), Ymark-(5)(4)) and the design position of the reference mark rm2 (Xmark_des2) (n), Ymark_des2(4)), has been stored in advance, and therefore, the processing of step S12 can be performed according to the design position of the storage towel. The processing in the above steps S11 and S12 can be in the same state as the above-described figure. + 乂 Align the microscope W2a of the optical system 16〇 with the imaging reference mark RM1' to photograph the reference mark RM2 with the microscope i62b (step su). In this case, the exposure area ER1 (measurement exposure area of the first item - (3) S3 200907596 reference marks RM1 and RM2 can be photographed. ~ Secondly, the photographed material is stored in the video financial institution (not shown), wrong shirt Image processing is performed by a processing mechanism (not shown), and the images of RMi and RM2 are taken to calculate the position of the ground (Xmark_l(l), Ymarkl(1)) and the base, RM2 ( The position of the X-those k_2(1), Ymark_2(1)), and then the reference mark RM1 of the area eR1 and the position of the reference mark 2 are stored in a storage mechanism (step S14). The position of the reference mark at this time (Xmark_l) (l), Ymark—丨(1)) and the position of the reference period RM2 (X job k-2(1), γ mine k_2(1)) are also the same as the piece by piece, which is the position of the image, for example, as long as it is The pixel can be positioned in units of a unit, and it is not necessary to use a unit of length generally used for millimeters, etc. The position of the reference mark paste of the exposure region ER1 obtained by the above method (ΧΠ1 blows 1 (1)&gt; blows 1 (1) )) &amp; SI mark RM2 (position of Xmark_2(1), Ymark_2(1))] is a bit coordinate in the X direction) and Y The position of the direction (γ coordinate). By the processing of steps SU to S14 described above, it is possible to detect the position X of the exposure mark ER1 (the measurement exposure area of the first item (n=1)), the position of the reference mark x, k-UDJmarkj(1)), and the position of the reference mark (2) ( Xmark — 2 (l), Ymark — 2 (1)) ' Then store the location. Next, it is determined whether or not the two reference marks RM1 and cafés of all the measurement exposure areas (ERi, coffee, ER9, and coffee 2) have been processed (step S15). When it is judged as NO, the processing of the above steps si2 to sw is performed again, and the next measurement exposure region is processed (step si6). Thereby, 84 200907596, when n=2, the position of the reference mark coffee (Xmark-1 (2), Y bribe! (7)) and the position of the reference mark (10) of the exposure area ER4 (Xmark-2 (2) is detected. ), the position of Ymark_2(2)), and then store the position at which the reference mark of the exposure area is set, and then the bit η is set (Xmark-1(3), Ymark-1(3)) and the reference is found. ^ Location of RM2 (Xlan k-2(3), Ymark_2(3)), then store ^ position ^ When n=4', determine the position of the reference mark of the exposure area ER9 (Xmark_l(4), Ymark_l (4)) and the base product &gt; RM2 (Xmark_2 (4), Ymark_2 (4)) position, and then store it: set. . If it is determined in step S15 that all of the measured exposure areas _, ER4, ER9, and ER12 have been processed, the alignment optical system 160 is moved away from the measurement position and moved to the retracted position (step s 1 7). By the above processing, the positions of the reference marks 4 of the four exposure areas Cong, the current 4, the ER9, and the ER12 (X ray k (4), Ymarkj (8)) and the position of the reference mark 2 (Xmark 1-2 (n) can be detected. ), Ymark_2(n)) ' Then save its location. As described above, in the step of aligning the alignment optical system 16G from the retreat position to the measurement position, the alignment optical M 16G system is maintained in the position of being positioned at the 敎 position [until all the exposure regions are processed, therefore, In step S12, the exposure region of the nth item is positioned in the illuminating field ea jiji, and the same state as in Fig. 5 (8) can be obtained, and the reference mark gamma is directly photographed by the microscope. The reference mark reading is taken by the microscope 162b. If the whole method is not adopted, it is not necessary to repeatedly perform the operation of positioning the alignment optical system (10) at the retreat position and the measurement position, thereby saving the time required for processing 85 200907596. Further, in the above-described example, the positions of the reference marks RM1 and RM2 are determined in accordance with the order of the exposure fields ER1 to &gt; ER4 - ER14 - ER9. By this, it is possible to shorten the moving distance of the X-direction moving Δ, the π 杉 moving stage 172 and the γ-direction moving stage ι74, and it is possible to carry out the time required for the processing of the step 七. j Reference mark (10) 1 and coordinate conversion of the position of the coffee > The four measurement exposure areas (10) obtained by the above processing, eR4, coffee, and _, the positions of the respective reference marks RMI (X job k-1 (n ), Ymark_1(n)) and the position of the reference mark with 2 (Xma(2(n), Ymark-2(n)), _ is set at the position of the coordinates of the image taken by the microscope coffee and (4). A, in order to calculate the correction position of the X-direction moving stage 172 and the γ-direction moving stage 174, it is necessary to convert it into a coordinate system fixed to the fixed disk center and mb. The conversion of the coordinate system will be described below. The exposure area EA, which is mounted on the exposure area ER1, ER4, ER9, or ER14 of the exposure substrate 156c of the stage 177, is positioned in the illuminable range EA, and the alignment optical system 160 is disposed. At the position of the measurement, the microscope 丨62a is used to capture the reference mark RM] by the microscope 162b. The reference mark RM2 is photographed by the microscope 162b. In Fig. 10, the upper side of the figure is also referred to as the χ direction. The left side is referred to as the +γ direction. As described above, the exposure area ER1 is referred to as the measurement exposure area of the item The exposure area ER4 is referred to as the measurement exposure area of the second item, the exposure area ER9 is referred to as the measurement exposure area of the third item, and the exposure area 丑14 is referred to as the measurement exposure of the fourth item of S6 200907596, and is indicated therein. In the case of measuring the exposure area, the term "the area to be measured by the nth item is referred to as an integer of η 卜 4. Here, the originally predetermined position ' is formed by the reference mark 于 in the measurement exposure. At the center of the left side of the area, the fiducial mark is read at the approximate center of the right side of the measurement exposure area. However, in the example of the figure ι, the reference marks RM1 and RM2 are both formed at the original position. As shown in Fig. 10, when the measurement exposure area of the first shot is positioned in the illuminable range EA, the position of the X-direction moving stage 172 in the x-direction is set to Xstg (4); In the position of the orientation, it is set to YSDg (npxStg(8), and the origin of the X-direction moving stage m is used as the origin, and the position in the X direction in the (C) and the (4) system is fixed. Coordinates. Similarly, Ystg(4) When the starting point of the γ-direction moving stage 174 is used as the origin, the position in the γ direction (γ-coordinates xStg(n) and Ystg(8)' in the coordinate system of the slave discs (10) and 178b is to be positioned in the χ direction. The amount of the movement stage 1 υ and the Υ direction movement stage 174 is predetermined. As described above, the two exposure areas er1, er4, er9, and Xie 2 of the exposure substrate B6c are two reference marks RM1. And the coffee, its design position has been: pre-storage. According to the two reference marks 镰...M2 design bit vw, 179 position, xstg(n) and sgn ' can be calculated and then stored in the direction of movement The stage 17 is directed to the control mechanism of the movement stage 174. Since the driving control of the y-direction moving stage m 87 200907596 and the * directional moving stage 1 74 is performed by a pulse signal, XStg(n) and Ystg(n) are compared with the number of pulses of the pulse signal. j. Here, Xstg(4) and Ystg(n) here are converted into an amount in units of actual length such as millimeters. Further, as shown in Fig. 10, the center of the measurement exposure region of the nth term is ER·0. Χ, the photographic area that can be photographed by the microscope 162a of the alignment optical system 160 is set to LSA, and the photographic area that can be photographed by the alignment optical "16Q _(10) 162b is set to 嶋. The center of the photographic area is LSA-O, The center of the photographing region RSA is RSA_〇. In the measurement exposure region of the nth item, the center LPS of the LSA is photographed in the coordinate system of the exposure substrate 156e by measuring the middle myocardial ridge of the exposure region as the origin. In the χ direction position (χ coordinate), the center ER_〇 of the exposure area is measured as the origin in the coordinate system fixed to the exposure substrate 156c, and the center LSA_〇 of the photographic area LSA is in the Y direction (Y The coordinate is YCCD-Bu-the same as the origin ER_〇 of the exposure area is measured, and is fixed in the coordinate system of the exposure substrate 156e, and the center RS A-0 of the photographic area is in the X-direction position (χ) The coordinates are XCCD-2. Further, the R-Ο of the exposure area is measured as the origin, and is fixed in the coordinate system of the exposure substrate We, and the center RSA_〇 of the photographing area RSA is in the γ direction position (γ coordinate). It is YCCD-2. By this, to measure exposure The center of the area ER_〇 is used as the origin, in the solid; t in the seat (4) of the exposed substrate core, the coordinate center of the photographing area LSA 'LSA-〇 coordinate system Xccd-uccd-i, the center of the photographing area RSA RSA·0 Coordinate system (XCCD-2, YCCD-2). Further: 88 200907596 As described above, the measurement position of the alignment optical system 160 is such that the reference mark RM1 can be photographed by the microscope 162a and the reference mark can be photographed by the microscope i62b. RM2 is the positioning position of the alignment optical system 16A, and the measurement position is continuously at a certain position. Therefore, the coordinates of the center LSA-0 of the photographing area lsa (XCCD_1YCCD-2) and the center RSA of the photographing area rsa The coordinates of -O (XCCD_2, YCCD-2), as long as the measurement position of the alignment optical system 160 is fixed, that is, the coordinate with consistency definition. ', and its coordinates (xccd-Uccd-2) and coordinates ( XCCD—2, YCCD_2), the analog sample is converted into units of actual length such as millimeters. Further, the position of the reference mark RM1 (Xmark—(10)) and the reference mark RM2 (Xmark_2(n), Ymark_2(n)) are as described above. Take the image of pixels (pixels) on the image as a unit The position 'however' in the coordinate system is converted into an actual nutrient unit such as a millimeter, such as the position of the reference mark RM1 and the position of the reference mark Wei2. For example, the table 177 can be set with high precision as a reference. Scale (not shown), by aligning: Learn the system to measure the scale, you can get the relationship between the number of elements and the actual long-term production. Using this relationship, you can change from the coordinates of the pixel to the actual length; The coordinates of the coordinates. &lt; Specifically, the χ coordinate fixed to the coordinate system of the photographing area 4 LSA is the Xmark-1 (n) of the reference mark in the X direction, with the center lsa_〇 of the photographing area LSA as the origin. In addition, the center LSA-0 of the photographing area ^ is used as the origin, and is fixed to the photographing area [the Y coordinate in the seat of the skull, and the reference mark is paid by the Y mark-1 (η) 〇89 200907596 in the γ # direction, The X coordinate of the coordinate system fixed to the photographing area RSA is the center point RSA-O of the photographing area RSA, and is the position Xmark_2(n) of the reference mark RM2 in the X direction. Further, the Y coordinate of the coordinate system of the photographing region RSA is fixed to the center of the photographing region RSA as the origin, and is at the position Ymark_2(n) of the reference mark RM2 in the Y direction. As defined above, when the position (Xmark_l (n), Ymark_l (η)) ' of the reference mark RM1 is converted from the coordinate system fixed to the image to the coordinate system fixed to the fixed plates 178a and 178b, the conversion type can be For: 'XM-l(n), YM-1 (ton

Xstg (8) + XCCD-1 + Xmark_l (8), Ystg(n) + YCCD_1 + Ymark_l(n)y (1) Further, the position of the reference mark RM2 (Xmark_2(n), Ymark_2(n)) is fixed from When the coordinate of the image is converted into a coordinate system fixed to the fixed plates 178a and 178b, the conversion type can be: 'XM-2 (8)) f, YM_2 (8) J \

Xstg(n) + XCCD_2 + Xmark_2(n), Ystg (8) + YCCD-2 + Ymark-2 (nL (2) Further, in the entire embodiment of the present embodiment, the reference mark RM1 and the reference mark RM2 are used. Midpoint position. The midpoint position of this reference mark RM1 and the reference mark RM2 can be calculated by: &quot;XM (8), '[XM-1 (η) + XM-2(n)]/2, [YM_1 (8) + YM—2 (8)]/2, (3) The variables used in the above equations (1) to (3) are also used to select the variable when measuring the exposure region in the above 90 200907596. n, any integer in the formula 4. By using the above formulas (1) to (3), the midpoint position of the reference mark surface and the reference mark RM2 can be rotated from the image coordinate system fixed to the image: The coordinate system fixed to the fixed plates 178a and 178b. The above-described coordinate conversion processing is carried out in the step si8 of the processing sequence of Fig. 9. Further, in the entire embodiment of the present embodiment, the design position of the reference mark RM1 is used. And the position of the design position of the reference mark rM2. The design position of the reference mark RM1 and the setting of the reference mark RM2 The midpoint of the position can be calculated as follows: XM-D (8), 'Xmark_desl (9) + Xmark_des2(n)]/2, (ton^Ymar^des^n) + Ymark_des2(n)]/2y ...(4) <Calculation of Six Parameters Sx, Sy, 0, ω, Οχ, and Oy> After the above coordinate conversion, (XM(n), YM(n)), and (4) obtained by the equation (3) can be used. (XM-D(n), YMjD(n)), using the following two formulas of the least squares method to calculate six parameters Sx, Sy, 0, ω, Ox, and 〇y: 91 ... ( 5) 200907596 ZXM_D(n) Συμ—d (8) Σι , r 1 + Sx , X -Sy χ (θ + τπ) ), 0χ &gt; ΣΧΜ_ϋ(η) ZYM_D(n) Σι '1 + Sx, X SyxG , Oy , (6) Σ XM—D (8)2 Σ XNUXn) X YM—D (8) Σ XM—D(8) x YM—D(8) Σ ΥΙ^η)2 ΣΧΜ—D (8) ΣΥΜ一D(n) &quot;ΣΧΜ (8) xXM_D (8), =ΣΥΜ (8) xYM_D@) ΣΥΜ(η) and Σ XM—D(n)2 SxM_D(n)xYM_D(n) YM_D(n) ΣΥΜ D(8)2 Σ XM_D(n) Σ YM_D(n) lYN^n)xXMJXn), =ΣΥΜ (8) xYM—D (8) ΣΥΜ(η) The variable η used in the above formulas (5) and (6) is also used to indicate the above-described variable of the measured exposure region. Aspect of system administration η ~ 4 in Shu any integer.

In the equations (5) and (6), Sx indicates the degree of expansion and contraction in the x-direction, as shown in Fig. 1 (1), and the expansion and contraction of the exposure region in the ±χ direction is expressed by this degree. This Sx corresponds to the magnification in the X direction. Sy indicates the degree of expansion and contraction of the Υ direction. As shown in Fig. 11 (b), the expansion and contraction deformation of the exposure region in the ±Y direction is expressed by this degree. This corresponds to the magnification in the above γ direction. 6» indicates the degree of rotation, as shown in Figure n(c), and the pattern of the exposure area is expressed by this degree. ω represents the degree of orthogonality, as shown in Fig. 1(d). The shear deformation of the exposed region is expressed by this degree. 〇x indicates the degree of displacement in the direction of X 92 200907596. The displacement of the figure is based on this degree... The first &amp; field is toward the ±x direction. 〇 y γ direction As shown in Fig. U(f), the degree of displacement of the exposure is as indicated. The displacement in the direction of the direction of the field is based on the processing of step 819 of this degree. The equations (5) and (6) and the method of 0" are used to calculate six parameters Sx, sy, θ, ω, 0χ. Γ 其 其 其 θ θ θ θ θ θ θ θ θ θ θ θ 取 取 取 取 取 取 取 取 取 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向 向For the batch, use the coordinate correction with the six parameters. ^ Use the stage m... to move: Use the two (4) X-direction shifting position of the universal moving stage 174: ' Sx ~Sxx(e + ®))pCM_D_) "〇x , one Λ - + , Syx0

Sy + = XM_le (m) YMJXmJ [OyJIvMje (4) (7) The m in the formula (7) indicates the variables of the i2 exposure areas ER1 to ER12 of the exposure substrate 156A. Here, m=1, indicating that the exposure area is paper...2 indicates that the exposure area is moving (four) indicates that the exposure area is YES; and that the exposure area is Talk. Also, ... indicates that the exposure area is deleted; (4) indicates that the exposure area is 跋7, = 7 indicates that the exposure area is (4) (4) the exposure area is ER5; m=9 indicates that the exposure area is café. Further, the claw ... indicates that the exposure area is ER10; 1 indicates that the exposure area is ER11; and m = 12 indicates that the exposure area is YES. Furthermore, in the following, when the exposure area of any of them is to be expressed, it is referred to as the exposure area of the nth item. Here, 93 200907596 m is any integer from 1 to 12. The xM_D(m) of the above formula (7) is the χ coordinate before the coordinate conversion, and is the 方向 direction position of the midpoint of the design position of the reference mark RM1 and the reference mark rm2. Further, YM_D(m) is the γ coordinate before the coordinate conversion, and is the γ direction position of the midpoint of the design position of the reference mark RM1 and the reference mark RM2. By calculating XM_le(m) and YM_le(m) using the above equation (7), the error of the linear component (hereinafter referred to as linearity error) can be corrected by the 丄-order equation, and the y-direction moving stage 172 and the γ direction can be moved. By moving the stage 174 to (XM_le(m), YM-le(m)), the exposure area of the mth item can be positioned at an appropriate position. Further, when there is no error, XMje(m)_XM_D(;mR γΜ−le(m)=YM−D(m), the x-direction moving stage 172贞Y-direction moving stage 174, It is positioned at the originally designed design position. <Transfer of pattern> Hereinafter, the pattern transfer formed on the reticle 142 is separately illustrated by the flow chart of Fig. 9 by the whole film method to the exposure shown in Fig. 8. The transfer order of the exposure areas ER1 to ER12 of the substrate (4). By using the formula (7), XM-le(1) and YMje(i) are calculated (step s2〇). Next, the stage 172 for moving the χ direction and the 丫 direction are driven. The moving stage a# moves the stage 177 to be positioned at the TMJe(1) and the top Je(1) calculated by the step S2(step). By this processing, the exposure area of the (4) item can be positioned after the correction. Then, the light emitted from the light source 110 is used to transfer the pattern formed on the reticle 94 200907596 (4) to the ER1 of the exposure substrate 156 (step S22). It is judged that the different exposure & P (step S23), when the discrimination process is not complete, then return to the above step S2, 俾 push &gt; When the result of the discrimination is shown in the next exposure area (step S24), the 12 exposure areas ε are patterned and transferred, and the processing of the whole mode is ended. 々 has been as shown in the whole method In the middle, the movement action of the wind, and the movement of the heart to the first position, the movement to the position of the test, does not need to be carried out in the complex transfer area, therefore, will be formed in the reticle &quot; When the substrate l56c is exposed, the processing time is expected to be shortened. Further, 'in the whole film type, W, , # | m storage and subtraction>' is used to determine the position of the reference mark 2 = shorten the time required for detection. The position of the fiducial mark of the four exposure areas in Fig. 8 is not limited thereto, and the position of the nucleus is also embossed. Thus, τ 1 is a reference mark A 172 of the exposure area. And the hunting average effect 'improves the X-direction moving load = the correction accuracy of the position of the moving stage 174. The position of the film is selected as the reference mark near the four sides of the selected exposure substrate; Hunt this 'rotation' parameter θ and the X-direction expansion parameter Sx, i:: The number sy' is not easily affected by the linearity error generated when the fiducial mark is formed. It is difficult to check even if the exposed substrate is damaged during manufacturing, etc. #+# % In this case, the position of the mark can also be used to estimate the position of the test result, and the stable 95 200907596 will be formed on the pattern of the reticle 142 for transfer. Order, exposure - ER2, 3H (four) - er7H mobile system in the order of Xie - ER1hER2. If you move to the adjacent exposure area as shown, you can shorten the X direction first. The time required for the transfer of the stage 174 in the Y direction and the Y direction is 172. The processing of the whole of the stencils <The error correction using the reference substrate> The positive film method has a good effect on the position of the base substrate formed on the exposed substrate due to the exposure of the exposed substrate itself... The displacement occurs from the original position. There is a "72, Y square; ^ Bay error" is generated by the movement of the load σ 172 in the X direction, the stage 174, or the structure of the 7th, and you can only use it for one degree. The stage 176 b can not be separated from the charging accuracy by the above-mentioned whole piece method, and the pattern cannot be rotated. The field A + only # + 丨王 exposes the substrate. Therefore, the reference substrate is moved in the x direction. Correction is performed by positioning the two shifting units 174. The reference substrate 158a is shown in Fig. 12(a)' is a schematic view of the reference substrate 1A entering its abundance 1... An enlarged view of the reference mark of one area. In the example of Fig. 12, the outer rectangle is a line segment representing the outline of the base. Further, the inner plate 158a is four and the longitudinal direction is three squares, There is a horizontal direction. 彳 is a square of 12 (ER1 ~ ERI2), their respective tables One area. The ® 甘甘&gt;- 佐 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 In order to clearly correspond to the position and size of the exposure region of the exposure substrate i56 (l56a, i56b' 1 56c), in FIG. 12(a), the exposure substrate 156 (156 &amp; 15 讣, 156 勹) The regions corresponding to the exposure regions ER1 to ER12 are assigned the same reference numerals. The reference substrate 158a' has substantially the same size as the exposure substrate 156. The reference substrate 1 58a is composed of a glass substrate. In the same manner, in the 12 regions ER1 to ER12, a plurality of reference marks are formed by chrome. In Fig. 12(b), the reference marks are indicated by white balls. Each reference mark has an error of 丨 micron. The reference substrate 158a is formed in the following manner. That is, when the reference mark is formed on the reference substrate 158a, the reference mark is within a range of i micrometers based on a predetermined position. The plurality of reference marks shown in Fig. 12(b) 'No need to decide its full The position of each of them can be selected by using a plurality of them. As shown in Fig. 13 (a), for each of the twelve regions ER1 to eri2, 18 reference marks are selected by a plurality of reference marks to be used. In the present embodiment, as shown in Fig. 13(a), the two reference marks of the black ball are paired, and the reference marks are regarded as (one reference mark pair, and nine reference mark pairs are added. RRM1 to RRM9 are regarded as a single basis. The nine reference mark pairs RRM1 to RRM9 are located near the center of each of the 12 areas ER1 to ER12, the vicinity of the four turns, and the midpoint of the four sides. Fig. 13 (b)' shows all the reference mark pairs of the reference substrate 158a. There are 12 areas ER1 to ER12 in the reference substrate 158a, and there are nine reference mark pairs in each of the 12 areas 97 200907596 ER1 to ER12. Therefore, there are 108 reference marks on the entire reference substrate. Furthermore, in Fig. 13(b), the nine reference mark pairs are each represented by i black squares. Hereinafter, the design position of the reference mark used for the reference substrate 158 &amp; is stored first. Specifically, the design position (x_k_des_L(k), Ymark_des_L(k)) of the reference mark on the left side of the reference mark pair, and the design position of the reference mark on the right side (Xmark_des_R(k), Ymark_des —R(k)) is stored first. Further, k = 1 to 1 〇 8, and as will be described later, k is used to select a variable of the reference standard pair. The χ-direction moving stage 172 and the Υ-direction moving stage 174 system (10) are moved in accordance with the design position of the reference mark.卩 由 由 〈 〈 〈 〈 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准 基准The variable k shown in the flowchart of FIG. 14 is used to select a specific reference mark pair from among the reference mark pairs of the figure = the reference area ER1 to ERU of the sensible substrate 158 &amp; For _~_ there are 1〇8, the value of the variable “. For example, if k=1, the table r represents the region 2=phan of the reference mark pair RRM1;] the reference mark pair RRM5. At first, the alignment optics is made. The system 16 is clamped at the measurement position (step Μ". : : Set and move the fixed or measured position, and the retreat position of the system is the same as the alignment optical "Μ". The better one, + the starting point of the first son system l60. In addition to ensuring that the photographer can be photographed by Gu Microscope, the position is also as long as the position can be 98 200907596 by the microscope 162a to the reference mark The reference mark on the left side of the median is photographed, and the reference mark can be photographed by the microscope 162b on the right side of the reference mark. Next, the drive substrate loading table 17&quot; the direction shifting stage 172 Moving the stage 174 in the 2 direction to move the table 177 to The first reference mark pair of the reference substrate 158a of the mounting table 177 is positioned at the photographable position (step S32). Next, the reference mark of the center on the left side is photographed by the microscope i62a of the alignment optical system 160 for the reference $ By the microscope i62&quot; the reference mark is used to photograph the reference mark on the right side of the reference mark (step call. - Then, the captured image is stored in the image storage mechanism (not shown), and the image processing mechanism (not shown) To perform image processing, take out the reference mark = pair: the image of the reference mark on the left and right sides to calculate the base rotation: the position of the reference mark on the left side (Xmark_L(k), Ymark_L(k)), And the position of the reference mark on the right side is H(k) 'Ymai> k-R〇〇)' and then stored in a storage not shown (step S34). The position of the reference mark pair at this time is on the image. The position, for example, the position of the image coordinate system in units of pixels (pixels) or the like may be used, and is not required to be positioned in units of lengths such as millimeters. In the above manner: the position of the reference mark pair of the waiter X direction Position (X seat) and position in the Y direction (Y coordinate) /, By the implementation of the above steps S32 to S34, the positions of all the reference marks RRM1 to RRM9 in the areas ER1 to ER12 can be determined, and then the position can be stored. 99 200907596 Secondly, it is judged whether or not p _ π + finds the oblique RRM1 $ has all the benchmarks for all regions ER ER12:, ~RRM9 has determined its position (step S35). If it is judged that the detection is not complete, 贞彳At the same time, the process of the above step S32 is performed to process the pair of grounds (step S36). The result of the discrimination of FRkt right in step S35 indicates that all of the reference mark pairs RRM1 to RRM9 in the 12 regions Κ12 have determined their positions. At this time, the alignment optical system 16()_opens the measurement position and moves to the retreat position ( Step S 3 7).

Further, in the above-described example, it is preferable to determine the position of the reference mark pair in the order of the areas eri - EM - ER4 - ER8 - &gt; ER7. - ER5 - ER9 - ER10. - ER12. As a result, the moving distance of the moving stage m#Y direction moving stage m can be shortened, and the time required for processing can be further shortened. <Coordinate conversion> Next, the following formulas (11) to 〇3) similar to the above formulas (1) to (3) are used to convert from the coordinate system fixed to the image to be fixed to the fixed plates 1783 and 17 Coordinate system. Furthermore, as described above, the position of the reference mark on the left side of the reference mark (Xmark_L(k), Ymark_L(k)), and the position of the reference mark on the right side (Xmark_R(k), Ymark_R( k)) is a position in the image coordinate system in which pixels (pixels) or the like are used as the unit, but in the following, the positions are converted into units of actual length such as millimeters. 100 200907596 XM—L(k), YM_L(kX : 'Xstg(k) + XCCD—L + Xmark_L(k), ^Ystg(k) + YCCDL + Ymark_L(k)/ ...(11) XM_R(k), YM—R(k), _ f Xstg(k) + XCCD_R + Xmark_R(k)' &quot;[Ystg(k) + YCCD_R + Ymark_R(k)y ...(12) XMJVl(k)] YM_M(k) J _ &quot;[XM_L(k) + XM_R(k)]/2&quot;| ~ JYM__L(k) + YM_R(k)]/2J (13) Xstg(k) in equations (11) and (12) And Ystg(k) are the same as equations (1) and (2), and are used to position the X-direction moving stage 172 and the γ-direction moving stage 174 in 12 areas ER1 to ER12, respectively. The XCCD_L and YCCD-L in the equation (11) are the same as the XCCD 1 and YCCD_1 of the equation (1). The center position of the imaging region of the microscope 1 62a. The XCCD-L is divided into 12 regions ER1~ The center of each of the ER12 as the origin 'in the X-direction position (χ coordinate) 'YCCD_L fixed in the coordinate system of the reference substrate 1 58a is the γ-direction position (γ-coordinate) in the same coordinate system. YCCD_R is the same as XCCD_2 and YCCD-2 of equation (1). It is the center position of the imaging area of microscope i62b. The XCCD_R ' The center of each of the 12 regions ER1 to ER12 is used as an origin, and the position (χ coordinate) 'YCCD-R in the coordinate direction fixed in the coordinate system of the reference substrate 158a is the γ-direction position (γ-coordinate) in the same coordinate system. According to the above formula (13), the point position (XM_M(k), YM_M(k)) of the reference mark on the left side of the reference mark pair and the reference mark on the right side can be calculated. In the following, the midpoint is used. The position is the position of the reference mark pair. 101 200907596 In the embodiment of the present invention, when the reference substrate 158a is used, the reference mark pair is used in the middle reference mark on the left side and the reference mark on the right side. The position of the point. The reference mark on the left side and the midpoint of the reference mark on the right side can be calculated as follows:

&lt;XRM_D(k), / , YRM_D(k)y V

[Xmark-des-L(k) + Xmark-des_R(k)]/2) [Ymark-des-L(k) + Ymark-des-R(k)]/2 j (14) <Nonlinear error Takeout > Next, using (XM_M(k), YM_M(k)) obtained by equation (13) and (XRM_D(k), YRM_D(k)) obtained by equation (14), use the following formula (15) and (16), the six parameters Sx, Sy, θ, ω, Οχ, and Oy are calculated by the least square method (step S39). r ZxRM_D(k)2 Σ XRM_D(k) x YRM_D(k) ExRM_D(k)N 'l + Sx , Σ XRM_D(k) x YRM_D(k) ZYRM_D(k)2 EYRM D(k) X -Sy χ (Θ + ¢5) , ZxRM_D(k) ZYRMJD(k) Σι , Ox , f Σ XM_M(k) χ XRM_D(k) Σ XM_M(k) x YRM_D(k) ZXM_M(k) \ (15) and 102 200907596

r ZXRM_〇(k)2 Σ XRM_D(k) x YRM_D(k) ZXRM_D(k)^ '1 + Sx, ZxRM_D(k) x YRM_D(k) ZYRM_D(k)2 ZYRM D(k) X Syx0 EXRM _〇(k) Zyrm_d〇c) Σι 1 Oy J Σ YM_M(k) χ XRM_D(k) Σ YM_M(k) x YRM_D(k) ZXM_M(k) (16) In the above formula (1 5) and (1) 6) The k used, as described above, is also used to select the variable for each of the 108 reference mark pairs. In the equations (15) and (16), Sx indicates the degree of expansion and contraction in the X direction, Sy indicates the degree of expansion and contraction in the γ direction, Θ indicates the degree of rotation, ω indicates the degree of orthogonality, and 〇χ indicates the degree of displacement in the χ direction. , 〇 y represents the degree of displacement in the γ direction. The linearity error can be taken by calculating the six parameters Sx, Sy, 0, ω, 0χ, and 〇y. Next, an error composed of nonlinear components (hereinafter referred to as nonlinear error) (XRM_nle(k), YRM_nle(k)) is calculated using the following equation: 'Sx, Sy χ θ - Sxx(0 + ©) YXRM_D(k)

Sy person YRM_D(k)

'XRM' e(k)) , YRMJe(k)J ...(17) (18) ^XRM_nle(k)&gt;| ^ f XRM_le(k) - XM_M(k)NL YRM-nle(k) J a URM_le (k) - YM: M(k), by using equation (17)', the approximation can be performed in one time using the six parameters calculated in the process of step S39, and the fiducial mark after correcting the linear error can be calculated accordingly The midpoint position of the pair. By using equation (18), the detection position of the midpoint of the nonlinear error can be subtracted from the position of the corrected linear 103 200907596 error f, and then used in (10) (mail curry - nle (k) )]:: Line: ^ 斤 取得 向 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动 移动S41;).···························································· In the structure of the crucible 6, the second-stage moving stage 172 &lt; ^ is generated. When driving Χ c or moving to the stage 174, the nonlinearity stored in step S41 can be determined as follows: 有 When there is an Abbe error, the 172 is the target and the direction is moved. Positioning of the loading stage 174 for carrying the 2 and Y directions. <Reference Substrate 158b> Fig. 15(a) is a schematic front view of the reference substrate 158b, and Fig. 15(b) is an enlarged view of a reference mark formed in the &quot;solid area of the reference substrate (10). In the example of Fig. 15 (4), the outer rectangle is a line segment indicating the outline of the reference substrate 8b. χ, on the inner side of the outer rectangle, there are four squares (four) to two legs (four) which are four in total and four in the longitudinal direction, each of which represents one area. Further, the reference substrate 158b is also the same as the reference substrate 58a, and is not a substrate for transferring a pattern. The square dotted line EB for indicating a region in the hook is a hypothetical line segment for exposing the substrate. The position and size of the exposure area of 156 (156a, 156b, 156c) have a corresponding relationship. In addition, in Fig. 15 (a), the exposed area of the light substrate 156 (1 56a, 156b, 156c) is exposed to 104 200907596. The regions corresponding to ER1 to ER12 are also given the same reference numerals. The reference substrate 158b' has substantially the same size as the exposure substrate 156. The reference substrate 15 8b is composed of a glass substrate, and further, in order to be described later. The reference pattern is formed on the reference substrate 158b by the reference pattern formed on the reference position by the reticle, and the sensitizer is applied to the reference substrate 158b. As shown in FIG. 15(b), the eri to ER12 are respectively in the 12 regions. A plurality of reference marks are formed. In Fig. 5(b), the reference marks are indicated by a white square and a black square. Each of the reference marks is formed on the reference substrate 158b with an error of 1 μm or less. which is The reference mark is formed at the position of the reference substrate 158b, and the position is within a range of 1 micrometer or less based on the predetermined position. It is not necessary to determine the name of all the plurality of reference marks shown in FIG. 15(b). It can be used in a few places, and the black square in Fig. 15 (8) indicates the reference mark used. As shown in Fig. 15 (4), there are 9 reference marks for each of the 12 regions ER1 to ER12. Therefore, the reference mark is used as the reference mark. m is used for 1〇; Fig. 16(a)' shows that the reference position of the reticle formed at the reference position is formed of a glass substrate by the reticle, and the reference pattern is composed of White =: sub-divided: covered by chrome. In the reference pattern, the white "the portion of the line" indicates the non-covered portion 44 in a state where the glass substrate is not provided. The covering system is maintained. <The reference pattern is on the reference substrate. 15 Transfer processing> 105 200907596 Figure 17 shows the reference pattern to the base, and the transfer processing of the reference A board 158h &amp; quasi-substrate M8b. *Base #substrate (10) used for the reference mark, its design Broken pre-storage. According to The reference mark moves the field and the leaf position to move the X-direction moving stage! 72 and the Y-direction move the door to move the wind from, 戟α174' can be aligned with the light, and the charge is 1 60 to mark the reference In addition, the number J of the figure shown in the flow chart of Fig. 17 is used to select one of the 1 to 8 reference marks not shown in Fig. 15(a). The number of the base soap is determined by the number of all the regions ER1 to ER12 of the reference substrate 158b J, and the total number of the variables jw is 108. For example, if j = 1, it indicates the reference mark IRM1 of the area (4); if j = 50, it indicates the reference mark IRM5 of the area ER5. First, the reference position is arranged with the reticle to the reticle (4) (4) shown in the figure, and the reference substrate is loaded on the table 177 of the substrate loading table 17 (step S51). Next, the y-direction moving stage 172 and the Υ direction moving stage 174 of the substrate loading table 170 are driven to move the table 177 so that the i-th reference mark of the reference substrate 158b of the table 177 is loaded. It is placed at a position where photography is possible (step S52). The positioning method of the reference substrate ι 581 in the pattern transfer (the pattern corresponds to the non-covered portion 44) is such that the center of each of IRM1 to IRM9 formed on the reference substrate 158b is located at the non-covered portion 44. center. In this state, the light for exposure is emitted by the light source, and the exposure light can pass through the non-covered portion 44 of the reticle for the reference position. The exposure light of the non-covered portion 44 is irradiated onto the reference substrate 15, and the pattern corresponding to the non-covered portion 44 is transferred (step S53). 106 200907596 Next, it is judged whether or not p ~ 疋 已 已 所有 IR ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER ER The processing of the next reference mark is carried out (step S55). By the above processing, as in the case of 1 - .rt Fig. 16(b), the pattern corresponding to the non-covered pattern is (10) (10) &amp; 44 The area ER1 to ER12 H The earth plate 158b has a right 疋 X-direction moving stage 172 or a moving stage 174 which does not generate a σ non-stationary movement error. In this case, the transfer side of ORM1 to ORM9 The smashing and smashing method is such that each of φ, t + 谷 〜 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 The center of the non-covered portion 44 corresponds to the center. The right-side direction moving stage 172 4Y-direction moving stage 174 and the raw movement error 'in this case' are as shown in Fig. 16(c). When the center position is shifted, the lower transfer patterns ORM1 to ORM9 are shovel. In this case, it is measured. The shift in the direction of the direction of the direction of the reference mark 16 is shown in Fig. 17 (4). The position detection of the reference mark pair of the reference substrate 158b will be described below using the flowchart of Fig. 17 . As shown in FIG. 16(b), the IRM1 to IRM9 are combined with the corresponding pattern 〇 〇 RM9 and are referred to as reference marks rrmi to rrm9. For example, 1 is 1 and 0 is 1 In addition, the variable shown in the flowchart of FIG. η is as described above, and the reference mark of the ray is selected, and all the reference marks rrm1 to rrm9 in the regions ER1 to ER12 of the reference substrate (4) have a total of 107 200907596. 1 〇 8 reference marks, the value of the variable j is, for example, the reference mark RRM1 of the display area ER1; J, the clothing reference mark surface 5. The leaf 5 〇, indicating the area ER5 =, causes the alignment optical system 16 to leave The off position is at the measurement position (step S56). The alignment optical system: set: the measurement position is the same as the piece by piece method. The position of the retreat position = the origin of the level optical system, the system 160. Further, the measurement position As long as the second _ comes to the benchmark mark issh &a Mp; stand up. Again, when using the base plate 158b, it is not necessary to simultaneously detect two reference marks, and one of the microscopes 162a or 162b will take a picture of i °. In the X-direction moving stage 172 of the disk substrate loading table 170, the Y-direction moving stage 174 is moved so that the first item reference mark pair of the α: reference board 158a is set, and the photographing position is set (step S57). ). Next, the reference mark is photographed by the alignment optical system (10) = mirror 162a or 162b (step 2 = item (4) is recorded in the X direction displacement x - dev (1) and γ direction shift X d is stored. At this time, the reference mark is in the χ direction The position of the bit etc. as the p shift - _ω ' can be set to the early position by, for example, the pixel (pixel) bit. It is not necessary to use the processing of the step 59 of the length sheet generally used for millimeters, etc. The 4M indications RRM1 to RRM9 calculate the χ-direction displacement 200907596 and the Y-direction displacement Y_dev(1), and the errors are stored. Next, it is judged whether all the fields of all the areas of the rhyme 12 = R rounds RRM9 are processed ( The step processing is complete, and the processing of the next reference mark pair is performed by the processing of the next reference mark pair (step S61). If it is the judgment result table ER1 in step S60, the processing is performed in the next step (step S61). The ER12 RR1U1 has a region U volume, and all of the RRM1 to RRM9 are processed to leave the alignment optical system 160 (step S62). After the heart position is moved and positioned at the retreat position, the step is performed by steps S38 to S41. The same processing as the above-mentioned nonlinear error is taken out The coordinates formed by the nonlinear error can also be calculated for the reference substrate (10), and the results can be graphically stored and stored. Further, in the flowchart of the above-mentioned ® 17, the processing of steps S38 to s4i is the same as that shown in FIG. 'There is the same symbol. <The pattern transfer process of the exposure substrate> As described above, the error of the reference mark due to the deformation of the exposed substrate, etc., U56a, i56b, 156c) can be obtained by the whole method. 2 to correct. Further, by using the reference substrate 158a or the ribs to form the pattern data, the nonlinear error can be extracted, and the nonlinearity error of the movement of the X-direction moving stage 1 Μ and the Y-direction moving stage 174 can be corrected. Therefore, when the pattern formed on the reticle 142 is transferred to the exposure substrate 156, the reference substrate 158a or 158b is used first to fabricate the pattern material (step S71 of Fig. 18). This processing is performed by the above-described Fig. 14 or Fig. 17 processing. 109 200907596 The whole piece will be formed on the reticle i42

Precursor base 156 (step milk of Figure 18). This processing can be implemented by the processing of Fig. 9. Further, in the case of the processing of L, the positioning position of the X-direction moving stage 172"2, step S12, is pre-made and processed in the processing of "S71 (four) (four) table 174. Way to store the location. Therefore, even if W moves to the moving stage 172 in both directions, an error occurs in the movement of m, and the x-direction moving stage &quot;2 and γ-direction movement can be performed without being affected by ? Use the stage! 74 positioning. Therefore, it is not necessary to use an interferometer or the like in order to determine the position of the x-direction moving stage m and the γ-direction moving stage 174, that is, the position=moving stage 17 can be moved in the direction of the stage m. In the case where the exposure substrate 156 is deformed or the like, and the position of the reference mark is displaced from the originally designed design position, the pattern can be surely transferred to the desired position. <The weighting process of the slice-by-chip method and the whole-chip method> With the implementation of the process of FIG. 18 described above, the position of the reference mark is generated by the deformation of the exposure substrate (lJ6a, 156b, 156c). The difference is generated, or the positioning error occurs in the x-direction moving stage 172 and the γ-direction movement = σ 1 74, and the position correction can be performed to transfer the pattern formed on the cobalt wire sheet 142 to the exposure substrate 156. . — However, as described above, the reference mark formed on the exposure substrate 150 is formed by melting the surface of the exposure substrate 156 by laser light, or by drilling, boring, or mechanical processing. Due to the accuracy of the formation of the reference marks 110 200907596, it is also possible that the reference mark is formed at the post-displacement position away from the originally designed position. In such cases, corrections must also be made for errors. Further, the printed circuit board is formed of glass fiber or the like, and it is not uncommon for a nonlinear error to occur in each step. In the following, the nonlinearity error of such an exposure substrate, together with the method of determining the exposure position, is explained below. Hereinafter, the exposure substrate 156a is used to use the difference between the position of the two reference marks RM1 and RM2 formed on the exposure substrate i56a, the position of the two reference marks RM1 and the position of RM2. The same processing as that of step S18 of Fig. 9 is performed, and the midpoint positions (XM(n), YM(n)) of the reference mark and the reference mark RM2 are obtained using the above equations (1) to (3). Further, the difference Δ M η (η) between the difference ΔXM(n) in the X direction and the Y direction can be calculated from the position data of the same. As described above, the position of the reference mark RM1 is ΧΜ-1(η), ΥΜ_1(η), and the position of the reference mark rm2 is (XM_2(n), YM-2(n)). Therefore, the difference Δ χΜ(η) is ΧΜ_1(η)−ΧΜ_2(η); the difference Δ ΥΜ(η) is ΥΜ_1(η)−ΥΜ_2(η). In the process shown in FIG. 9 described above, the exposure substrate 156 is used to process the detection of the reference marks RM1 and RM2, and is performed for the four exposure regions eri, ER4, ER9, and ER12, however, the exposure substrate is used. In the case of 156a, the reference marks RM1 and RM2 are detected for the twelve exposure areas ER1 to ER12. Therefore, the above-described variable n is an integer value of 1 to 12. In this case, 'n=1 is an exposure area of ER1; 11=4 series exposure area 111 200907596 is ER4 '· n = 5 system exposure area is ER8; n = 8 system exposure area is er5; n = 9 system exposure area is ER9; n=12 system exposure area is eri2. The detection of RM1 and RM2 is performed sequentially from n=丨 to [2. By the implementation of this process, the y-direction moving stage i72 and the Y-direction moving stage 174 are moved to the phase. The adjacent exposure area is sufficient, so that the moving distance of the stage can be shortened, and the processing time can be reduced. Furthermore, it is also possible to perform detection of a certain mark and RM2 for the 12 exposure areas ER1 to ERi2, but Benchmark only for the selected complex (4) exposure area Detecting RM1 and RM2. 实施Second! After performing the above-mentioned test of the reference mark and reading 2, the wrong reference to the reference substrate 158a is used first, and the non-line is taken out by the cat soil. The moving stage 172 and the "error to the movement of the movement" and then the X-direction movement carrier Α 72 72 ^ direction of the mobile load ... position in the corrected nonlinear error carrier, for the detection of the reference mark job and help. In the following, the position of the shift position in the X direction (7) and the position in the X direction (6) and the position 6 after the positioning of the table 174 are moved (X^or^^n #^i2 4 : Ί &amp; '(XC〇r(4), Yc〇r(4)) is used as an example. When the X-direction moving stage 172 is moved in the direction of the 丫, the system is in the exposure (4) ER4, and the corrected stage is The position 74 is set to obtain the midpoint position Μ(4) of the difference between the reference marks RM1 and RM2 (10) (4), YM (4)) and the χ direction, and then the processing shown in Fig. 19 is performed. () The difference in the Υ direction Δ 112 200907596 First, the reference mark RM1 is used. With the midpoint position of RM2 (ΧΜ(η), ΥΜ(η)), and the corrected stage position (X_cor(n), Y-cor(n)), Using the equations (21) and (22) of the least squares method, six parameters Sx, Sy, 0, ω, 〇x, and 〇y are calculated (step S81). f Σ X_cor(n)2 Σ X_c〇r(n ) x Y—cor(n) Σ X—cor(8) &quot;l + Sx ) Σ X_cor(n) χ Y_cor(n) Σ Y_cor(n)2 Σ Y_cor(n) χ -Syx(0 + tj) Σ X_c〇 r(n) Σ Y_cor(n) Σΐ ^ , Ox v J ZXM(n)xX_cor(n) Σ ΧΜ(η) χ Y_cor(n) Σχ一cor (8) Σ X_cor (8)2 Σ X_cor(n) x Y_cor(n) Σ X_c〇r(n), &quot;l + Sx, Σ X_cor(n) x Y_cor(n) Σ Y_c〇r (8)2 Σ X-cor (8) X Syx0 , Zx_cor(n) Συ一c〇r (6) Σΐ , °y &gt; Σ YM(n) x X_cor (8), Σ YM(n) x Y_cor(n) (21) (22) Σ Y_cor(n) After performing the processing of step S81, use equation (23) The approximate positions of the correction positions of the midpoints of the two reference marks RM1 and RM2, that is, the positions X_le(n) and Y_le(n) are calculated in the order of the equation (step S82). &quot;Sx, Syx0

Sx X (Θ + ω), ^X_cor (8), 'Ox, rX" e(n), Sy , , Y-cor (8) J , 〇 yJ , YJe(n) (23) As described above, the movement of 174 Here, X_cor(n) of the formula (23) and the direction shifting stage 172 and the Y-direction moving chapter 113

X 200907596 The position of the stage after the nonlinear error has been corrected. Next, using the equation (24), the nonlinear error components _1116(11) and Y_nle(n) are calculated for the midpoint positions of the two reference marks RM1 and RM2 (step S83). 'Xnle(n), 'XM(n)-X_le(n), and Y_nle@\ (24) Next, using equations (25) and (26), six parameters Sx, Xy, and 6 are calculated for the difference in position. ;, ω, Ox, and Oy (step S84). Σ X_cor(n)2 Σ X-cor(n) x Y-cor (8) Σ X_cor(n) '1 + Sx , Σ X_cor(n) x Y_cor(n) ΣΥ—cor (8)2 Συ cor(8) X -Sy χ (θ + ϊπ) Σχ cor(n) V _ Σ Y_cor(n) Σι / , 〇χ , 'ΣΑΧΜ(η)χΧ一cor(n), =Σ ΔΧΜ(η) χ Y_cor(n) (25) Σ X_cor( n) Σ X_cor(n)2 Σ X_cor(n) χ Y_cor(n) Σ X_cor(n) Σ X_cor(n) χ Y_cor(n) Σ Y_cor(n)2 Σ Y_cor(n) X SyxG Σ X—cor (9) Σ Y_cor(n) Σι, 〇y &gt; Σ ΑΥΜ(η) χ X_cor(n), =ΣΔΥΜ (9) xY_cor(n) ... (26) Σ Y_cor(n) Then use equation (27) to 1 The approximate positions ΔX_le(n) and ΔY_le(n) of the difference between the positions of the two reference marks RM1 and RM2 are calculated in the following equation (step S85). 114 ...(27) 200907596 &quot;Sx ,δγχθ

Sx X (θ + ω), 'X_cor (6), 〔Ox, , AX_le (8), sy , , Y_cor〇i)y 卞, ΔΥ_le(n), Δ X_le(n) calculated according to equation (27) Δ Y_le(n), the transmission equation (28)' differs from the difference between the positions of the two reference marks and rm2 by the nonlinear error components ΔX_nle(n) and ΔY_nle(n) (step S86). '△X—nle (8), '△ΧΜ(η)-ΔΧ—le (8), ΔΥ—nle(n)y, ΔΥΜ(8)-ΔΥχηχ (28) which is linear with respect to equation (27) The error components Δ xje(n) and Δ Y−le(n)′ are calculated by a statistical method to be 3 σ — le (step S87). Here, σ·le is the standard deviation of the linear error components Δ X —ie(n) and Δ γ—ie(n). In the following, the above 3 σ -le is set to S_le. Next, the nonlinear error components Δ X_nle(n) and Δ Y - nle(n) calculated by the equation (28) are calculated by the statistical method v (step S8S). Here σ-nle is the standard deviation of the nonlinear error components Δ X-nle(n) and Δ Y_nle(n). In the following, the above 3σ_nle is set to S_nle. Furthermore, 'S_nle' is used to indicate the error information of the magnitude of the nonlinear error component. Further, the error information 'the same as s_me calculated by the nonlinear error components X_nle(n) and Y_nle(n) calculated by the above steps may be used in the following steps. By using χ le(n) &amp;Y "(η) calculated by the processing of step SS2, 〇(11) and Y_nIe(n) calculated by the processing of step S83, - processing at step 115 200907596 S87 The calculated S_le, together with the S_nle calculated in the processing of step S88, calculates Xpos and YpOS by the equations (29) and (30) (step S89). Xpos = (SJex(X_le + X_nle) + S_nlexX_le)/ (S_le + S_nle) ... (2o\

Ypos = (SJe x (YJe + Y_nle) + S_nle x Y_le) / (S_le + S_nle) The X-direction moving stage 172 and the Y-direction moving stage 174 are positioned in the equations (29) and (30). ) Calculated XpOS and Yp〇s. The positions (x_ie, γ - le) in the above equations (29) and (30) are positions corrected by the entire chip method. On the other hand, the position (x"e + x_nie Y-le + Y-nle) ' is the position corrected by the slice-by-chip method. If the two reference marks RM1 and RM2 are excessively large when formed on the exposure substrate 156, (In addition, when the deformation of the exposure substrate 156 also includes a high order of three or more times, 'S-nle tends to be large. Furthermore, the second error of 156 is small, S_le=0. In this case, when the exposure substrate can be regarded as a random error, the equations (29) and (30) are shown below, and the pattern can be transferred to an appropriate position in a one-piece manner.

Xpos=X le... (31)

Ypos=Y_le··· (32) On the other hand, if the two reference marks rm 1 and RM2 have a small error when formed on the exposure substrate 156 of 116 200907596, S-nie=〇. Further, when the secondary error of the exposure substrate 156 is large, SJe is superior to a large value. The error in this case can be regarded as the second-order distortion error of the exposure substrate 156, and the equations (29) and (3) are expressed as follows, and the pattern can be transferred to an appropriate position in a sheet-by-piece manner. (33) ...(34)

Xp〇s=X_le + x nie

Ypos=Y_le+Y nie * Further, although both S-nle and S-le are small values, the nose position is obtained by the equations (29) and (30). Furthermore, when calculating equations (29) and (30), the coefficient D can be used to make s_nle = s-nlexDe coefficient D, which is used to bias the shirt to the whole piece or to the piece by piece method. The person is ^ ^ 疋 』 is determined by the user's meaning. For example, if it is set to D &gt; 1砗, the 志-甘μ^ 旰 table is not biased to the whole piece;

When V is set to D &lt; 1, it means that it is biased in a piece by piece manner. <Replacement of Error> In the above-described processing, the difference between the difference A XM(n) in the X direction and the difference J in the γ direction, J knives ΔYM(n), is performed in Fig. 19 . Since the two reference marks RM1 and RM2 are arranged along the χ and Χ directions, the difference Δ ΧΜ(η) in the X direction corresponds to S λ, the degree of expansion and contraction of the universal direction, and the difference Δ YM in the Y direction. (n), corresponding to the table, the degree of clothing is not red. Therefore, after the processing of the above-described step S84, the Q n is processed to calculate the six parameters Sx, Sy, ω, Ox, and 〇y, the X can be replaced with the difference in the X direction. 117 200907596 Δ XM(8) Further, the Θ can be changed to the difference Δ ΥΜ (η) in the γ direction to perform the processing shown in Fig. 19 . Further, it is not necessary to replace both Sx and Θ, and only one of them may be replaced. The position correcting mechanism of the first working position, the position correcting mechanism of the crucible 2, the reference substrate position control mechanism, the reference substrate position storage mechanism, and the reference substrate exposure mechanism can be configured by the first control unit of the first embodiment. The reference substrate position correction operation, the exposure substrate position control mechanism, the exposure substrate position storage mechanism, the exposure substrate reference mark position storage mechanism, and the linear error correction calculation mechanism. <Intelligent Whole Film Method> The exposure substrate may be formed into a sheet which can be approximated by a second-order or a third-order method due to thermal deformation during manufacturing or a shearing force of a predetermined size, for example, In the case of 24(a)*, the deformation of the exposure substrate can be opened by the second-order method, or as shown in Fig. 24(b), and the deformation of the exposure substrate can be approximated by the third-order equation. The exposed substrate in the manufacturing process has a few deformed shirts as shown in Figures 24(4) and (b). Therefore, the position of the reference mark on the exposure substrate is also shifted to a position which can be approximated by the second order, the 3: minus type, or the like. Since the displacement of the position is due to the error of the position of the reference mark, the error of the displacement corresponding to the reference mark can also be approximated by the second-order method, the third-order method, or the like. In general, the deformation of the shape of the exposed substrate is in many cases where the positional error can be approximated by the order of the cubic equation. In the following description, the deformation of the shape of the exposure substrate is approximated by a third term (including a second term). On the other hand, the reference mark on the exposure substrate 56 is formed by exposing the surface of the substrate 1 56 by laser light 118 200907596 or by mechanical processing with a drill or the like. Therefore, it is possible to shift the reference mark to a position different from the originally intended design position due to the accuracy in forming the reference mark. The displacement is a random displacement and a random error occurs at the position of the fiducial marker. Further, such a random error in the position of the reference mark can be regarded as an error which can be approximated by the term of the fourth-order or more, as compared with the error which can be approximated by the third-order expression. In the present embodiment, the error occurring in the position of the reference mark is divided into two types, namely, an error caused by the deformation of the exposure substrate and an error caused by the formation of the reference mark, and the former is divided into two types. It is regarded as an error that can be approximated by the third-order equation (the third-order approximation error); for the latter, it is treated as an error (random error) that can be approximated by the higher term of the fourth-order equation or more. In the smart overall method of the present embodiment, when the position error of the reference 桴圮 is corrected, the third-order approximation error or the random error is used as the dominant error, and the position of the substrate is exposed at the time of exposure. The best and corrective methods are different. Also, P, the error of the position of the reference mark is 2, and the magnitude of the approximation error and the random error are used as a reference to determine which is the dominant error, and according to the result, the above-mentioned piece by piece method is adopted. The same method as in the weighting process of the entire film method is used to correct the error of the exposure position. As described above, the correction of the position error of the exposure substrate during exposure is referred to as "smart whole film mode". Furthermore, the intelligent whole film method is not limited to the third-order type, and the mans: a 'right' approximates the position error and random error by the high-order error of the third-order (such as 4-time or 5-^). The size of the second criterion is used to correct the error in the exposure position during exposure, which is also acceptable. 119 200907596 According to the processing 1 of the above-described smart whole film method, the loading table of the substrate for carrying the exposure substrate during exposure can be more accurately positioned at the target position. Fig. 25 is a flow chart showing the processing sequence of the $hui type whole chip mode. The following is a detailed description of the smart whole film method according to the figure. In the example of the smart whole chip method shown below, the magnitude of the approximation error and the random error of the 次3 formula are used as the criterion for judgment. First, the reference mark (exposure substrate reference mark) of each exposure region on the exposure substrate is detected. This test is carried out in the same manner as described in Fig. 9, Fig. 17, or Fig. 17 (step SHH). ® 26 is an overview of the exposure substrate used in the processing of the smart whole film. The exposure substrate 156d' has four exposure areas (1) 丨) to er (4, sentences, etc. The exposure area is generally represented by a natural number of (10) (four) ~ 4:1 to 4). Each of the exposure regions has four exposure substrate reference marks ΚΜι (ί, 』), ΚΜ 2 (υ), then (10), and RM4 (i, j). Further, the detection of the exposure substrate reference mark in the present embodiment is performed by using two exposure substrate reference marks as the i group. When the position of the exposure substrate reference mark is detected, all of the exposure substrate reference marks of all the exposure areas can be sequentially detected. However, in order to shorten the detection time of the exposure substrate reference mark, the preferred method is to omit the detection of the partial exposure substrate reference mark. In this case, it is preferable to replace the position of the omitted exposure substrate reference mark with the position of the other exposure substrate reference mark in the vicinity thereof. In the following, the exposure substrate of Fig. 27 is added. In the exposure substrate 丨 56e, the exposure substrate reference mark for which the detection is omitted is not indicated, or is indicated by a broken line. As shown in the exposure 120 200907596 optical substrate 156e, the exposure substrate reference marks RM3 (1, 1) and RM4 (1, 1)' of the exposure region ER (1, 1) have been omitted from the detection process. The position of the exposure substrate reference mark is calculated based on the positions of the exposure substrate reference marks RM1 (2, 1) and RM2 (2, 1) in the vicinity thereof. Similarly, the detection of the exposure substrate fiducial mark for other exposure regions can be omitted in the exposure substrate 156e. As shown in Fig. 26, when detecting the exposure substrate reference mark of the exposure substrate i 56d, it is necessary to detect 32 sets (64) of exposure substrate reference marks, whereas the exposure substrate i56e shown in Fig. 27 is used. When the exposure substrate reference mark is detected, it is only necessary to perform detection of 20 sets (4 pieces), and the time required to detect the exposure substrate reference mark can be shortened. Further, in Fig. 26 and Fig. 27, the exposure reference marks to be detected are indicated by black squares. Fig. 28 shows an example in which the detection of the exposure substrate reference mark is further omitted. In the exposure substrate 156f, the detection of the exposure substrate reference marks RM1 (1, 2) and RM2 (1, 2) of the exposure regions ER (1, 2) is omitted. The position of the exposure substrate reference mark can be calculated from the positions of the exposure substrate reference mark RM2 (1, 1) of the exposure region ER (丨, 2) and the RM1 (1, 3) of the exposure region ER (1, 2). . The other exposed substrate reference marks are also the same, and the exposure of at least one of the exposed substrate reference marks is detected in the vicinity of the exposed substrate reference mark in the vicinity of the exposure area. The substrate reference mark can be omitted for the detection operation. As a result, in the case of exposing the substrate 156e, it is only necessary to detect 15 sets (30) of exposed substrate reference marks shown by black squares in the figure. As described above, by omitting a plurality of exposure substrate reference marks, 121 200907596 can shorten the base capacity, and can reduce the number of movements and stops of the alignment optical system, and the time required for the detection of the alignment mark. The &amp; can improve the exposure processing system in a preferred manner. θ, in addition, the reference mark of the partial exposure area may not be detected, but the detection of the reference mark may be performed only for the selected plurality of exposure areas. In this case, the position of the exposure region which is omitted is detected, and the approximate position can be calculated by interpolation based on the position of the exposure substrate reference mark of the vicinity of the exposure region. X ’ can detect the reference mark even if the exposure area is not used for some reason. Similarly, the approximate position of the exposure area can be calculated by interpolation based on the position of the exposure substrate reference mark in the vicinity of the exposure area. ^ Again, one exposure area does not necessarily require four exposure substrate reference marks. For example, as shown in Fig. 4 (4), in the case where two exposure substrate reference marks are provided in each exposure region, it can be handled by this method. However, in order to achieve a more accurate exposure position, the position of the exposure embossed plate fiducial mark is more precise. Preferably, each exposure area has at least 4 exposure substrate fiducial marks. After the processing of the above step S1〇1 is performed, the position of the exposure substrate reference mark is calculated from the detection result of the position of the exposure substrate reference mark (step S102). The meaning of "calculating the position of the exposure substrate reference mark" means that when the position detection of a part of the exposure substrate reference mark is omitted, the position of the omitted exposure substrate reference mark is calculated based on the position of the other exposure substrate reference mark in the vicinity thereof. . Moreover, the position of the substrate reference mark is exposed by using equations (1) to (2) (that is, from the coordinate system fixed to the image 122 200907596 t to the coordinate system fixed to the fixed plate _ and (10) ) The same method of %. Before the position of the exposure substrate reference mark is calculated, the reference substrate is preferably subjected to error correction as a pretreatment. In other words, it is preferable to use the same reference substrate as the reference substrate (10) or (10), and to correct the error caused by the structure of the \ directional movement stage 172 and the directional movement stage m in advance. The χ-direction moving stage (7) is corrected to the moving stage 174' after the error caused by the structure of the 移动-direction moving carrier ..."-direction moving stage 174 is corrected. (χ—c〇r(iJ), where 1 and J ' are natural numbers of 1 to 4, indicating the exposure area. For example, (X-corhj)), Y_CGr(i, j)) The χ-direction moving stage (7) 舁Y-direction moving stage 174 is positioned to the stage position after the correction of the exposure area ER(i,". ^Secondly, the entire substrate is over-exposed to the exposed substrate reference mark' The least square method of the position 'calculates the total area error parameter value, and the characteristic error value is obtained according to the displacement of the exposure substrate reference mark (step S1 〇 3), where the total area error parameter is expressed, The resulting overall error, specifically, is the same as the six parameters Sx, Sy, θ, ω, 〇χ', and 〇y obtained by the slice processing. The calculated position of the exposure substrate reference mark is the same as that of steps S18 and S19 of Fig. 9 Specifically, 'the same type as the formula (3) is used to obtain the paired benchmark; the midpoint position of the own. X, using the least squares method (21) and (10), (8) 123 200907596 Instead of η, the approximation is performed in a least squares method to calculate the full region error parameter. Next, for each of the detection areas, based on the position of at least two exposure substrate reference marks, the detection area error parameter value SxG of the error characteristic based on the displacement of the exposure substrate reference mark is calculated, 』), ^(1), θχ(υ }, = and Sy(i, j) (step S104). The detection area here refers to each area for separately calculating the detection area error parameter value. Further, although the detection area may be the same as the exposure area. Area, but in order to improve the accuracy of this process, each exposure area can be subdivided and the subdivided multiple areas can be used as detection areas. X 'depending on the situation' can also be used in size, position, shape and exposure area. The different area is used as the detection area. In the above-described example of the processing of the step coffee, the detection of the partial exposure substrate reference mark is omitted. However, the detection area in which the detection of the exposure substrate reference mark is omitted is also possible. The position of the exposure substrate reference mark of the omitted exposure area is calculated based on the position of the other exposure substrate reference mark located nearby. In the example, the exposure area and the detection area are °, so the symbol of the detection area is described as the detection area ER(i, j) using the same symbol as the exposure area. The detection area error calculated by the processing of the above step S104 The error characteristic of the parameter 2' is based on the displacement or deformation of each detection region, for example, the degree of expansion and contraction of each detection region in the direction of the root direction and the 丫 direction, or the degree of deflection or rotation of each detection region in the X direction and the Y direction. In the example of FIG. 29, the shape of the detection area after the detection and the deformation and the position of the exposure substrate reference mark 124 200907596 are indicated by a broken line and a solid line, respectively. In the middle, the positions of RMi(i,j) and RM3G,·)) are not displaced, but the position of RM4(i(1) is displaced. Dx and Dy are expressed in the X direction and γ after the deformation of the detection area. The distance between the center and the center of the two exposure substrate reference marks adjacent in the direction. The difference between the coordinate components in the X direction or the γ direction of the exposure substrate reference mark adjacent in the X direction or the γ direction is removed by Dx The length after Dy becomes and Sy(i,j)(2). The average value Sx(i,j) of Sx(i,j)(1) and (3)(7) can be obtained, and Sy(i,j) can be obtained. (i) an average value Sy(i,j)D with Sy(i,j)(2), and Sx(i,j) and Sy(i,j) are a detection region error parameter, which is indicated by The degree of error in the expansion or contraction in the X-direction or the γ-direction in the detection region ER(i,j). The difference between the coordinate components in the γ-direction or the X-direction of the exposure substrate reference mark adjacent in the X direction or the Υ direction, 0 x (i, j) (i), 0 x (i, j) (2), 0 y (i, j) (l), and 0 y (i, j) (2). X〇,j) is taken as the average of Θ x(i,j)(i) and β x(i,j)(2), and θ y(i,j) is obtained as θ y(i,j ) (i) and the average of 0 y(i,j)(2). The 0 x(i,j) and 0 y(i,j) are the errors of the detection area error parameter ', which is the shear force in the X direction or the Y direction in the detection area ER(i,j). The degree of deformation, or the degree of rotation when 0 χ (ί, 】) and Θ y (i, j) are the same size. In general, the detection region error parameters Sx(i,j), sy(i,j), 0 x(i,j), and 0 y(i,j) may be referenced by a plurality of exposure substrate reference marks in a certain detection region. The difference between the coordinates of the position is calculated. Therefore, the detection area error parameter value corresponds to the first derivative of the position of the exposure substrate reference mark s. The test 125 200907596 used to measure the area error parameter is not limited to Sx(iJ), Sy(iJ), 0, and Θ y(i,j) ' /, which is equivalent to the position of the exposure substrate reference mark. ^ The subdifferential and error characteristics are based on the displacement of the substrate reference mark, and are not limited to any type. In addition, when the exposure substrate is deformed and can be approximated by the second-order equation, the detection region error parameter is a value corresponding to the secondary coefficient of the second-order equation. Further, it is also possible to obtain the minimum flatness of the equations (5) and (6) based on the coordinates of the position of the exposure substrate reference mark included in the detection area ER(i, j) without using the difference method described above. In the method, the parameters sx〇, j), syo'm (6)), ω (6)), 〇χ (υ), and 〇州, 』) of each detection area are calculated, and then these are regarded as detection area error parameters. Then, in the respective detection areas, the linear component (step) of the detection region error parameter value is calculated according to the detection region error parameter value and using the least square method, that is, the midpoint position and detection for the detection region ER(i, j) = The domain error parameters Sx(i,j), SyW), Θ X(i,j), and 0 y(i,j) are implemented by the same process as step s 19 of FIG. 9 to obtain the detection region error parameter 曰, · Qi〗 〖Generation knife. Specifically, the description is the same as the description of step S 〇 3, and the midpoint position of the detection area ER(i, j) is obtained first. Secondly, instead of Δ YM(n), use the same formula as the least flat and 缶 (25) to (26), replacing ^:^]^11), Ν1') with 8?^"). Substituting (i, j) for n, the six parameters Sx, Sy, θ, ω, Ox, and Oy are calculated by the least square method: approximate method. x is the same as the method of equation (27) 'I can calculate the error parameter of the detection area. The knives sx~le(1j)&amp; Sy-le(i, u. and 126 200907596 are the same processing, and In the case of Neng Zeng, the linearity of the error parameter value of the 检测 b b 检测 区域 ' ' ' ' ' ' ' ' ' 算出 算出 算出 算出 算出 算出 算出 ' 算出 算出 ' ' 算出 算出 算出 算出 ' ' ' ' ' ' ' ' ' ' Here, the "1st-order difference knife" is the difference between the error value of the detection area of the two adjacent detection areas, and the "2nd order difference" means the difference of the error value of the detection area by 13⁄4. The difference is further calculated. Similarly, when the difference is calculated n times, the difference is referred to as "n-order difference". Further, as described in the above steps s 1 〇 4, in general, The detection area error mic is calculated based on the difference between the position coordinates of the exposure substrate reference mark. Therefore, the detection area error parameter corresponds to the i-order difference (丨 differential) of the position of the exposure substrate reference 。. 1st order difference of the regional error parameter value (1 The differential is also equivalent to the second difference (secondary differential) of the position of the substrate reference mark. Generally, the difference of n steps (n-differential) of the detection area error parameter value is equivalent to the exposure substrate reference mark. η+1 difference of position (η+1 subdifferential). In the following case, the first-order difference Δ SxQj) and Δ θ of the detection area error parameter value are calculated for only two adjacent detection areas. ), and △ (9 y(ij). Secondly, based on the first-order difference and using the least squares method, calculate the differential linear component Δ Sx(i,j)Je, △ Sy(i,j)Je, △ Θ X(i j)Je and Δ 0 yG'j)-le (step S107). The calculation of the difference linear component using the least square method is the same as the case of calculating the linear component of the regional error parameter value in the above step S105, and can be performed using the same equations as in the equations (25) to (27). 127 200907596 As described above, the detection area error parameter corresponds to the first differential of the position of the exposure substrate reference mark. Therefore, the difference of the detection area error parameter value corresponds to the second derivative of the position of the exposure substrate reference mark. Therefore, when the shape of the substrate can be approximated by the third-order equation due to the deformation of the exposure substrate, the difference linear component is a value corresponding to the third-order coefficient of the third-order equation. Next, for each detection region, according to Sx ( The detection region error parameter value such as i, j}, the linearity of the detection region error parameter value such as Sx_le(i, j), and the cumulative sum of the differential linear components of ΔSx(i, j)Je, etc., and the detection region is calculated. Error information of the error parameter value Sx_nle(iJ), etc., S_Sy, S_θ x, and S jy (step sl8). For example, the error of the detection region error parameter value of &amp; i Sx-nle (iJ) is calculated by subtracting the cumulative component of the differential linear component Δ δ χ(Μ) je from the detected region error parameter value Sx(i,j) by the linear component SxJe(i) added to the error value of the detection region error parameter , j} is known as Sx-le-totaihj. That is, Sx_nle(i, j) can be calculated by the equations (35) and (3ό).

Sx_le-totalG, 』) = Sx_le(i, j) + (△SxRj) - cumulative sum of u) (35) Sx_nle(i,j) = Sx(i,j) - Sx_le_total(i,j) . ..(36) and 6· y_nle(i,j), can also be Sy_le_t〇tal(i,j), Q and then be the same as equation (36) for Sy_nle(ij), x-nle(ij), and the same The equations of X"e-total(i,j) and Θy_le_total(i,j) are calculated by the equation (35). 128 200907596 It can be understood from the above calculation process that the linearity of the detection region error parameter value &gt; Sx_le(i, j) is equivalent to the differential differentiation of the position. Further, the differential linear component or the like corresponds to the second derivative of the position, and the cumulative sum of the differential linear components corresponds to the integral value. Therefore, after the cumulative sum of the differential linear components is added to the linear component of the error value of the detection region, the obtained Sx_le_totaiQj) corresponds to the third-order approximation error in the error according to the displacement. Further, since the accumulation and the equivalent are equivalent, when the differential linear component can be expressed in a mathematical expression, the integral value of the equation can be used as the cumulative sum. On the other hand, the error Sx of the detection region error parameter value is such that, in order to remove the third-order approximation error from the total error of the displacement, it is considered to be equivalent to the random error. Further, in the above-described step S1 ,, when the difference between the two adjacent detection regions is different from the second-order difference value of the detection region error parameter value, when the cumulative sum of the difference = linear component ΔSxGJ)_le is calculated The cumulative sum of the times must be calculated from the order of the difference. For example, when calculating the difference of the second order, the cumulative sum of the difference linear components of the second order is calculated, and the cumulative sum of the cumulative sums is further calculated, and then the cumulative sum of the cumulative sums is added to Sx_le_total(i , j). Similarly, when calculating the difference of the nth order of the third order or more, the cumulative and repeated steps are performed n times in the same step, and then added to t〇tal(i, j). The error of the detection area decision parameter value can be calculated based on the Sx. For Δ Sy(i,j)Je, Δ Θ x(i, jUe, and Δ «9 y(i,j)_le are also the same. X, then 'calculate the error information S_Sx, S_Sy, S 0 129 by statistical method. 200907596 and S_ Θ y, which are equivalent to the unevenness of the detection area error. For example, the error Sx-nle of the test = ^ parameter value (the standard deviation of i (4), according to the ^ domain The error information of the error of the error parameter value may be 2 times or 3 times (3σ) or the like as the error of the error value of the detection area error parameter. For example, 3' calculated by s, etc. As the error information of the error of the detection area error parameter value n, the 3σ calculated by Sy_nle(i,j), etc., or θ yXM) can be regarded as the error of the error of the detection area error parameter value. Information S_Sy, S_θχ, or Sjy. Furthermore, in the error information, the inhomogeneity of the error equivalent to the error value of the detection area is not limited to the standard h as long as the error can be expressed statistically; For the sexes, various types can be used. Secondly, the error information S_S based on the error of the detection area error parameter value x S_Sy, S-0 0, and s_ 0 y, and a weighting coefficient w is calculated, which is used to indicate an error of the cumulative sum of the linear component and the differential linear component corresponding to the detection region error parameter value (equivalent to 3 The sub-approximation error) and the size ratio of the error (corresponding to the random error) that does not correspond to it (step S109). The meaning of the "size ratio" of the cloud here is that the error corresponding to the approximation error of the third-order equation The error corresponding to the error of the random error is used as a reference to understand the error of which is dominant. An example of the weighting coefficient is as follows: the error information of the error value of the detection region error parameter calculated in step S108 is the detection region. The error parameter value is used as a divisor to divide it, for example, using equation (37). 130 200907596 W-(S_Sx/Sx)x(S_Sy/Sy)x(s_0x/0x)x(s_ey/ey).. (3 7) The receiver's calculation of the target position to be positioned (step S110) based on the position of the exposure substrate reference mark, the total area error parameter value, and the weighting coefficient W (for example, 'if the equation (37) The calculated weighting factor w=〇, here The shape and the error of the position of the reference mark correspond to only the third-order approximation error, and the displacement of the substrate and the displacement of the exposure substrate reference mark accompanying the substrate are not random. Therefore, the position of the exposure substrate reference mark calculated in step S102 can be determined to be true. Here, when the weighting factor w=〇, the target position to be positioned on the loading table is calculated based on the position of the exposure substrate reference mark calculated in step S102. The target position calculated by the above method is set to (10), do not). The calculation method used in calculating the target position (Xd, Yd) can also be the same as the method of correcting the position by the above-described borrowing and filming method. Further, the target position (Xd, Yd) can be calculated from values corresponding to the cubic approximation errors of S, Sy_le_tQtaiw), ΘX-le-t〇taKi, D, and Θ y. On the other hand, if the weighted filament W calculated by the equation (37) is 1, in this case, the position error corresponding to the reference mark has only a random error. Therefore, the position of the (4) optical substrate reference mark calculated in (4) S1〇2 is not correct. Here, the total position parameter value calculated by (4) S1G3 can be used to calculate the target position when the loading stage is positioned to each exposure area, thereby using the loading stage for loading the exposed substrate during exposure, and the predetermined target position. It can be located in a more precise place. The position of the target position of 2009, 200907596 calculated by the above method is set to (Xg, Yg). In general, the weighting coefficient W is used to apply weighting to the above two methods of calculating the ^ ^ TTR ^ setting. For example, the target position (Xpos, Ypos) can be calculated using the following equations (38) and (39). Head -..(38)

Xpos = (1-W) X Xd + W X Xg

Ypos = (1-W) x Yd + W x Yg (39) Furthermore, the calculation method of the 'weighting coefficient and the target position is not limited to the above example. The significance of the weighting coefficient is that the coefficient can be 3 The sub-approximation error and the magnitude of the random error are used as a reference to determine which is the dominant 2 error' and then reflected in the calculation of the target position of the exposure substrate, and the calculation method can be appropriately changed according to the above meaning. Further, the weighting coefficient is not limited to one, and the weighting coefficients can be calculated one by one for each detection region error parameter value for weighting calculation. Next, the loading stage is positioned at the target position calculated as described above to transfer the pattern to the exposure substrate (step Sill). As described above, in the case of processing the smart whole mode, the third-order approximation error and the random error are used as the determination criteria. However, the predetermined number of times can be repeated in steps S106 and S107, and the appropriate calculation is performed in step sl7. The sum of the differences in the number of times makes the judgment criterion of the intelligent whole-chip processing become the position error and the random error of the exposure substrate reference mark caused by the deformation of the substrate of the third-order (four-time or five-time type, etc.). Large 132 200907596 Fig. 30 (4) and (8) show an example of correction of the position error when processed by the above-described smart whole chip method. The arrow in _ indicates the correction direction and the size of the school i obtained when processing in a smart whole chip mode. As is apparent from the above diagram, an excellent error correction result can be obtained when the deformation of the exposure substrate produces a positional error which can be approximated by the second-order &lt;3rd-order equation. Further, as shown in Fig. 3 (4), by the processing of the smart whole chip method, even if the change in the size of the rotation which can be approximated by the following equation occurs, the error can be corrected in this case as well. <Specific method> Next, explain the specific implementation of intelligent alignment. In the above, the differential value is obtained to obtain the second-order component and the third-order component of the error. However, when there is a case where the mark cannot be detected and an error occurs, or the exposure position is completely different from the mark position, it is necessary to The exposure position is obtained by interpolation based on the measurement result of the mark. The specific description is as follows. The size of the printed circuit board is about mmx5Qmm; when the projection lens is used, the exposure area is about 2 mm to about 3 mm, and the 遂 becomes the stepwise repeating exposure of the number of divisions 3x2 to 4x4 A right. In the conventional alignment optical system, the two methods shown in Fig. 31 were used. Figure 3 1 (a), in the case of two alignment sensors for the exposed area, in order to measure the black ball mark in the exposure area * with two alignment sensors, the upper side is measured first. After the marking, the stage is moved so that the lower target 2 is measured, and after the alignment optical system is retracted from the exposure area, the exposure position is right based on the measurement result to perform exposure. 133 200907596 Good to open another aspect - Figure 31(b) is a case where there are four alignment senses for the exposed area. The H line is marked with a mark of #4, so that the alignment optical system declares „禾术&amp; The exposure position is the advancement of the rod t. In either case, the advanced lamp is aligned with the movement of the sensor before the exposure of each illumination point. Firstly, the method described so far == the initial alignment of the alignment Finally, the use of its two quasi-sensing device for the whole-piece alignment, intelligent alignment method. Here, the ten Zhuo set 'in Figure 32 (a) is configured into a 4 ^-position full-piece alignment. Slice alignment, in Figure 32 (b) is another aspect of wisdom, when the customer has a higher φ Λ&gt; the requirements of the order of the order, as shown in Figure U(4) or Figure 32(d), Then there will be no mark; small gap, and the warehouse will be in the center of the field or between the areas of the narrower outside or the center of the wider k shape, without the need to insert the alignment sensor in this situation 33 ( b) will be aligned with bb = inner ' and can be placed outside the exposed area as shown in Figure 33 (4) or * a 4 . Although the sensor must be in accordance with the mark = side, although once, The drive is completely set up between the batches, but it can be a highly stable structure as long as it is set up. That is, when there is a drive detector, when a τ # ^ ^ 对准 alignment sensor is obtained, ...view, J (again, it is not uncommon for the average situation to drive sensing, .., and the duration of the mind to change). In the case of Fig. 33(a), the result of using a 2 difficulty to move the position of the a position to obtain a 16 point device is to perform 8 cases of 3 alignment sensing n. 33(b), using the movement of 5 positions to obtain 15 points 134 200907596 Load: The moving steps are as shown in Fig. 33 (4), Fig. 33 (4), and the moving distance of the moving moon is short, and the processing time is short. According to this, the measured % / ^ Bessie, from the mark on the periphery of the printed circuit board, set the position of the ten-seat's deviation from the error value. The distortion of the printed circuit board, the reason is that the material properties of the glass material are included, and since the system is continuously twisted, the moon b will be the heart. Self; the value is concatenated into a line, and the value in the middle can be predicted. Preferably, the interpolation is based on the interpolator of the curve according to its continuity. As in the above test, as long as there are four measurement results up to the third time. In turn, the curve γ = Αχ3 + Βχ2 + cx + D can be determined. Further, if the mark is set to a position such as Fig. 33 (a) or Fig. 33 (b), more marks can be set for measurement without depending on the exposure position. The more the number of markers, the more * can be used to approximate the delineation of the mark itself due to the averaging effect. The method of approximating the curve is quite a number of 'in the real' state, the system is averaged. A higher least square approximation is used to illustrate. / After taking y measurement points for the equation F(X) of the curve, when Τ = Σ (y - F(x)) 2 D), find the function coefficient (A, b, c, which minimizes the above T). System

F(x) ==A*x3 + b*x2 + c*x + D 135 200907596 This is the least square approximation to become a simultaneous equation. Therefore, it is considered to be the condition that the formula τ = becomes the minimum. J equation. T can be regarded as a function of the coefficient (A, (Y ' A * X3- B * T is the minimum condition for B, C, D), and the condition that T is the minimum is the differential of T = the equation of 〇 / 5 D = 〇占丁 / 5 c = 〇&lt;5 T/ 5 B = 〇占丁 / 5 A = 〇解上式 and then become a 4-ary simultaneous equation (the following formula). Σχ°, Σχ1, Σχ2, Σχ3, ί 4 Ν / η η, Σχ1, Σχ2, Σχ3, Σχ4 Λ η Σχ°γ ) Σχ2, Σχ3, Σχ4, Σχ5 D η Ex'y V Σχ3, Σχ4, Σχ5, Σχ6 匕, β J Σχ2γ / ^ Σχ3γ,

The determinant is solved to calculate the coefficients A, -A * X3 + η * 2 , D. In addition, the number - Α χ + Β * χ 2 + 〇 * χ + Γ λ ^ D is stored, and not only the measured values of the measured coordinates, but also the corrected values. The interpolation of the ❹ curve is shown in Fig. 34, which is used to describe the position of the position 34(a) of the approximate calculation of the curve 砗 θ (4), and the coordinate value of the 6 standard design on the printed circuit board side can be obtained. Pu 1 (X ρ U'y) ~ pu6 (x, y), with 6 markers measured 136 200907596

Qul (x, y) ~ Qu6 (x, y) (black ball). However, since the two alignment sensors are used, if the stability of the alignment sensor is unsatisfactory, the average of the two data can be taken to become the three marker design coordinates Pu 12 (x, y) ~ Pu5 6 (x, y), and three marker measurements Qul2 (x'y) ~ Qu56 (x, y) (cross shape). The description of the black ball and the cross shape is calculated in the same manner, and since it differs only in the number of measurements, the description will be based on black balls. In the same way, get 6 mark design coordinates on the underside of the printed circuit board? 41 (\, 丫) ~? (16 (\, 7), and 6 mark measurement values Qdl (x, y) ~ Qd6 (x, y) (black ball). On the other hand, three mark design coordinate values Pi1 can be obtained from the left side of the printed circuit board. (x, y) ~ P13 (x, y), and 3 note measurement values Qll (X, y) ~ Q13 (x, y) (black ball); from the right side can get 3 mark design coordinate value Prl ( x, y) ~ Pr6 (x, y) and three marker measurements Qrl (x, y) ~ Qr3 (x, y) (black sphere). The resulting (PU, QU), (Pd, Qd), (P1, Q1), (Pr, Qr) are divided into groups'. The coefficient β for calculating the equation for each group according to the above method. In the above determinant, Pu and Pd' are fixed values in the y direction. In the calculation of the determinant, Pul~pu6=xl~x6, Pdl~Pd6=xl~x6. In the case of Qu and Qd, since the X direction and the Y direction are the same variables, they become two approximations corresponding to the respective directions. Curves (QuX(X), QUy(x)), (Qdx(x), Qdy(x)). On the other hand, in the case of P pr, since the χ direction is a certain value, therefore, in the determinant In the calculation, P11~Pl6=yl~y6; Prl~Pr6=yl~y6. In the case of Qu and Qd, the X and Y directions are the same as the variables, and become the respective parties. To the corresponding two approximate curves 137 200907596 (Qlx(y), Qly(y)), (Qrx(y)), (Qry(y)). As described above, after solving the above determinant, Figure 34 is obtained ( c) and the curve data shown in Fig. 34(d). The curve data is based on the equation. For Pll and Pd, the interpolation value can be calculated anywhere on the coordinate axis. Also, for PI and Pr. The phrase 'as long as it is on the coordinate axis' can calculate the interpolation value anywhere. As shown, the above equation can calculate the value on the fixed axis of 1 column or 1 row. However, the coordinates on the plane of the exposure coordinate system are determined. The following steps must be taken to expose the coordinates. First, for the curve groups Qu and Qd that are opposite each other, and the curve groups Q1 and Qr that are orthogonal to each other, the exposure position coordinates (Εχ, ^) are respectively shown in the figure. Each of them is substituted as shown in Fig. 35(a) and Fig. 35(b). The value obtained by the upper group (Qux(Ex), Quy(Ex)) The value obtained by the lower group (Qdx(Ex), Qdy(Ex) The value obtained in the left group (Qlx(Ey), Qly(Ey)) The value obtained in the right group (Qrx(Ey)'Qry(Ey)) The value of the value is 'if the exposure coordinate position is on each curve In the case of shape, it indicates the degree to which the exposure position must be corrected by the position of the exposure coordinate. = The exposure position is close to the top, and the result will become the data of the upper group. The position of the light is close to the lower one, making it close to the lower group. The fruit of the information is: That is, if it is close to the upper group, it must be the data of the upper group: If it is close to the lower group, the data of the lower group must be weighted. As described above, 'for Pll, pd -, ., and s, since the y direction is a certain value, and the y coordinate is set to Puy, the γ coordinate of the lower group is set to kiss 138 200907596, (Puy-Ey) /(Puy-Pdy) : (Ey-Pdy)/(Puy-Pdy) becomes the internal division ratio of the exposure coordinates. Since the closer the internal points are, there is a larger weighting, and therefore, the correction value at the exposure coordinate position can be calculated by the following calculation formula. Correction value xl=(QUX(Ex) * (Ey-Pdy) + Qdx(Ex) * (Puy-Eyh/Ruy.pdy) Correction value yl=(QUy(Ex) * (Ey_Pdy) + Qdy(Ex) * ( Puy Ey)) / (Puy_pdy) This method, although only the correction of the exposure coordinate position (Ex, Ey) is known in advance, it is natural to expose, because of the size of the exposed area, the coordinate position of the light region of 4 隅 can also Here, as shown in Fig. 35 (hook and Fig. 35(d), if the four coordinates are substituted into the equation by the same method as described above, the correction value in the exposure region 4隅 can be obtained. Due to the overall exposure of the area, there is no high-order correction.

The linear error of the heart and the shot. The least square of the above-mentioned full-chip alignment is calculated as a positive value. It is possible to calculate the method of each of the above, and until now, only to use the data of the upper group 139 200907596 « The data with the lower group, however, as shown in Figure 35 (a), Figure 35 (b) It can be seen that the same correction value can be calculated by using the pair of the left side group and the right side group. As described above, in the case of P1 and pr, the χ direction is a certain value. Therefore, the X coordinate of the left group is set to ρ1 右, and the χ coordinate of the right group is set to Prx, then (Plx-Ex) / (pix_prx) : (Εχ·ρΓχ)/(ρ1χ_ρΓχ) is the ratio of the inside of the exposure coordinates. The correction value at the exposure coordinate position can be calculated by the following calculation formula.杈 positive value xl=(Qlx(Ey) * (Ex-prx) + Qrx(Ey) * (Pix_Ex))/(plx_prx) 杈 positive value yl=(Qly(Ey) * (EX-Prx) + Qry(Ey * (ρΐχ_Εχ))/(ρΐχ prx) However, when only one pair of data is used, as shown in Fig. 36 (Dian can't correct the error in the X direction twice or more. On the other hand, only use left and right In the case of a pair of data, as shown in Fig. 36 (b), it is not possible to correct the error of the y direction twice or more. In the embodiment, the method is to first use the data of the upper and lower 以 to determine the correction value. Secondly, from the data of the left and right pairs, the common term between the data of the upper and lower pairs, that is, the correction value of the deviation value and the order component, is removed, and only the correction value of more than 2 times is added. Any of the shapes of 36(a) and 36(b) can be corrected. However, when the measurement marks of the upper and lower and left and right data are different, or when the number of measurement points is large, there is also a deviation value. It is not possible to have a commonality with the 丨 component. In this case, the deviation from the data of both parties can also be taken out from the data of the two parties. By this method, the correction value of all the exposure positions located inside can be calculated by the measurement result of the mark formed on the outer circumference. Further, the above-mentioned system shows the approximate value of the third-order curve, but the third-order curve The approximation must have 4 points of data. Although there are 6 points for the upper and lower data, there are only 3 points for the left and right data. In this embodiment, when there are more than 4 points of data, the curve approximation is performed 3 times, but in the case of 3 points of the material line 2 times the curve near calculation, 纟 2 points can be used to calculate the sub-equation to the next person's calculation. That is, if the detection of 6 points of data, there are several points mark &quot; line If the error is measured, it can be eliminated, and the best is taken according to the remaining numbers: 』Intelligent whole-chip alignment is used to detect the mark corresponding to all the illumination points. Among them, there are 3 approximation errors and random errors. The differential is processed in a different way. The high-speed intelligent whole-chip pair of flags placed on the periphery is the same, and T = E (the value of T of YA*x3_B*x2_c*x_DhMa is above the predetermined value, which can be regarded as a mark. The possibility of depicting errors is ancient in this case, and there are also It may be deteriorated by the method of correction. In addition, although the above-mentioned measurement marks of the upper and lower left and right data are the same and the one-time component is not common, the stated person may be a matter. In this case, the measurement may be ignored. The number of points, and the value or deviation value is not the degree of the i-th order component, but the correction of the automatic ^ T approximation curve, or the correction method of the linear approximation. X, if ί 2 has been known beforehand, the substrate The distortion tendency φ 2 times curve can be corrected 疋 when 疋 can be corrected by straight line approximation ′ can also be determined by means of parameter input f 141 200907596 approximation method. On the contrary, in the case of mark drawing, the shovel production*, , and β' in the case of the smear drawing are better in the bit processing or the averaging effect of the slashing, and are drawn when the exposure of the mark is used. error. In addition, when the printing of the Thunder is not a singer | the occurrence of " Α * χ 3 Β * χ 2 road plate is also very small, Τ = Σ (Υ _ C XD) 2 value will be very small. In particular, in In the case of not having the third component, the coefficient A is close to 0, and when the component is not, the coefficient B is also close to 〇. If only two errors are caused by c and two, only the printed electricity must be obtained, and the linearity of the 疋The number of points is small, and the number of points can be ==, and the setting is also included. In the present embodiment, the second piece of the same batch is paired with the values of ',, and Β The number of measurement points of the quasi-marker is reduced. As described above, the error of the printed circuit board is determined by the approximation method in the group. As described above, if the group of the upper and lower groups is widened, the same description can be applied to the group: The statement is as follows. For the upper and lower groups, &lt;linear approximation average value&gt;

Bu ; Qd = Ad * (χ) + ua v V Au (χ) + 正 00 heart to repair the average of the first component

Qu =( Au + Ad)/2 * (X) + Bu Qd =( Au -t- Ad)/2 * (X) + Bd 142 200907596 •• Offset value X, deviation value Y, rotation, orthogonality,倍〇平X, 俾 rate γ (same component as the whole slice alignment) &lt;Line Approximation&gt;

The first-order Q vu ^ Au * fx\ is calculated by a linear approximation in a pair of groups.

Bu ; Qd = Ad * (χ) + Bd. Deviation value X, deviation value Y, rotation, orthogonality, magnification γ, trapezoidal X, trapezoid γ " ^ &lt; 2nd curve approximation &gt; 2 pairs of curves are approximated by 2 times in a pair of pairs, . , Qu = Au2 * X) + U * (χ) + Cu ; Qd = A(j2 * (χ) + Bd * (χ) + μ. Deviation value χ, deviation value γ, rotation, orthogonality, Magnification 梯形, trapezoidal X, trapezoidal γ' ϋ shape x'u shape γ, shape Υ, pincushion X, pincushion Y s ; X, cylinder &lt;3rd curve approximation&gt; 3 times in 1 pair The curve approximates the third-order formula J;;AU3MX) + Bu2*^ + CW + -^-Ad3,^ d (x) + Cd * (χ) + ) . Phase-engraved value X, deviation value Y, rotation, orthogonal Degree, 仵 rate 梯形, trapezoidal X, trapezoidal Y, U-shaped χ, υ-shaped γ, 倍:”, Pincushion X, Pincushion Υ, 3 times curve error X,: 2:, tube 3: Undergrowth error Χ 3 times magnification error γ curve error V, the method used in the above-mentioned 'group', the most straight line, the approximate line, the second approximation curve, the most: the flat method, but the method Not limited to the (four) line of the square is limited to this 'can also be moved by the spline (spUne) function or 143 200907596 The average is obtained and so on. Furthermore, it is also possible to insert data into a line such as a straight line that is bent into a line, and to interpolate the line (10) of a line or a curve. Furthermore, it is also possible to find the fixed number two to process, and to pattern the exposure position data. In this case, the law is used to unify the vocabulary, and the similar information is used as a description. The curve furnace of the service is as shown in Figure 32 (4), and there is also a parallel with the side in the middle part. In the case of the group, the curve obtained by the likes and the thousands of points is obtained. Λ The straight line or near “curve lean machine” uses the internal information to determine the exposure position. Moreover, by using the fine γ 5丨丨古站山 on the lower side of the middle set plate, the curve is obtained by straight line or approximation:: using the internal information to determine the exposure position, this is also possible. Example. It is a matter of course that the present invention is divided into two, and the present embodiment is divided into two, and the present embodiment is used twice. ...If the above is handled by the smart whole chip method, even if there is a reason why the exposure substrate reference mark cannot be detected in the exposure area for some reason' (4), the load table is calculated in each exposure area using the total area error parameter value. The target location at the time. Moreover, even if there is a case where a part of the exposed substrate reference mark cannot be detected, the position of the exposure substrate reference mark which is detectable at the periphery thereof can be used, and the position of the undetectable exposure substrate reference 'α has been approximated. Interpolated. Therefore, exposure can be performed in the same manner in the exposure region where the exposure substrate reference mark cannot be detected for any reason. By the above-described processing of the smart whole film method, the exposure position for exposure 'has taken into account the random error of the position of the exposure substrate reference mark. 144 200907596 • Also, since the detection of a part of the exposure substrate reference mark can be omitted, the throughput of the exposure processing can be improved. Further, even if there is a case where the exposure substrate reference mark cannot be detected for any reason in the exposure region, the exposure processing can be performed in this case as well, and there is a more accurate exposure processing. Furthermore, there are many cases where the error of the stage structure is not a random error. Therefore, it is also possible to use a reference substrate to correct errors due to the structure of the stage, etc. In addition to the error correction of the exposure substrate, it is also possible to perform correction for errors due to the stage structure or the like. That is, the use of the reference substrate to correct errors due to the stage structure or the like may be omitted herein. <Second Embodiment>> FIG. 20 is a schematic view of a projection exposure apparatus 2 according to a second embodiment of the present invention. The projection exposure device is the same as the projection exposure device i 1 (10), and is mainly used for manufacturing a printed circuit board. The projection exposure apparatus 200 is a projection exposure apparatus using a digital micromirror element r CDlgltaljvilcr〇mirror_Deviee), which is a method of transferring a pattern to an exposure substrate without using a reticle. <Projection Optical System> The projection optical system includes a light source 210, an optical fiber 212, and an optical rod 214. The light source 2 1 〇 has a plurality of ultraviolet LED light sources (not shown). The light beam emitted from the optical source 210 is incident on the optical fiber si?, and is concentrated by the optical fiber si] into the optical beam, and then the illumination beam is emitted from the emitting portion of the optical fiber 212. The light beam emitted from the fiber is incident on the optical rod 214, and the optical rod 214 causes the illumination of the incident illumination beam to be nearly uniformized and then emitted. 145 200907596 <Digital Micromirror Element Optical System> An illumination relay optical system 222, a deflection mirror 224, and a digital micromirror element 230 are disposed in the traveling direction of the light beam emitted from the optical rod 214. The light beam emitted from the optical rod 214 is adjusted to the size of the reflecting surface 232 of the digital micromirror device 230 by the illumination relay optical system 222 and the deflecting mirror 224, and then irradiated to the digital micromirror device 230. The reflecting surface 232 of the digital micromirror element 23 is composed of a plurality of mirrors that can be independently driven. By driving a plurality of mirrors, the angle of the plurality of mirrors can be varied to reflect light incident on the plurality of mirrors to a desired direction. The plurality of mirrors are arranged in a two-dimensional manner, for example, 800x600, 1280x1024, and 1980x1080. <Enlarged illumination system> An enlarged illumination system 240 is disposed below the digital micromirror device 230. The magnified illumination system 24A includes a plurality of lenses. In the magnified illumination: the optical path in the system 240 has a 曈 position, and a shutter 242 is disposed at the 瞳 position to form an opening 244 in the shutter 242. As described above, the plurality of mirrors of the digital micromirror element 230 can change the angle. In the present embodiment, the angles of the plurality of mirrors of the digital micromirror device 23 are driven to be the first angle or the second angle. In the first angle, the beam incident on the plurality of mirrors is directed toward the shutter 242 (four) σ 244. When the angle of the plurality of mirrors becomes the first angle, the light beam emitted by the light source 210 will pass through the opening 2 4 4 of the shutter 242 and will fall downward toward the lower side. On the other hand, the second angle The beam of the plurality of mirrors is incident on the opening 244 which is not the shutter 242. When the angle of the plurality of mirrors becomes the second angle, the light beam emitted by the light source 2 10 will The mask 242 is shielded from passing through the opening 244. Below the magnified illumination system 240, a microarray lens 250 is disposed. The microarray lens 250' is comprised of a plurality of mirrors each of which is coupled to the digital micromirror element 23. Each of the plurality of microlenses is formed into a corresponding relationship. When the angle of the plurality of mirrors becomes the first angle, the light beam emitted by the light source 21 会 passes through the opening 244 of the shutter 242 and is incident on the microarray lens. 250. The microarray lens 250 converges light incident on each of the microlenses to form a beam spot at a position LS below the microarray lens 250. The beam spot has a size of several micrometers to ten micrometers. <Projection lens 150 Quasi-optical system i 6 〇, substrate loading table 7 〇 > Below the microarray lens 250, a projection lens 丨 5 〇, an alignment optical system 1 60, and a substrate loading table i 7 配置 are disposed. Projection lens 丨 5 〇 The alignment optical system 160 and the substrate loading table i 7 have the same configurations and functions as those of the first embodiment of the invention, and are denoted by the same reference numerals. The table i of the substrate loading table 170 77. The exposure substrate 156 is mounted in the same manner as in the jth embodiment. Further, the exposure substrate 156 may be any one of the exposure substrates 156a, 156b, or i56c described in the Jth embodiment. The beam spot image formed at the position LS is formed. Since the beam spot image is formed on the exposure substrate 156, the spot where the beam spot image is formed can be exposed. Therefore, by using each of the plurality of mirrors The angle is controlled to any of the above-mentioned 肖1 度 degree or 帛2 度 degree 147 200907596, so that the surface position of the exposure substrate 156 is made to be somewhat like where the formation and the dot formation are formed. Since the exposure substrate 156 can be formed The image formation The place where the person is not formed is equal to the transfer of the pattern and the pattern. Fig. 21' is an enlarged view of the beam spot image formed on the exposure substrate 156 and the exposure substrate 156. A part of the beam spot image 260 formed on the exposure substrate and a portion of the exposure substrate 156. In Fig. 21, the beam spot image 26 表示 is indicated by a white ball. The image is formed on the exposure substrate as shown in Fig. 21 The beam spot image of i 56 is separated from each other by the position of the adjacent image. Therefore, the position exposed by the exposure base is also discretely distributed, and it is difficult to form a continuous conductor pattern of the printed circuit board on the exposure substrate 15 . ^ Based on the above-described mirror elements, as shown in FIG. 21, the arrangement direction of the light beams is slightly inclined (angle Θ t) from the moving direction of the exposure substrate 156. Further, in Fig. 21, the exposure substrate 156 is moved in a direction in which the white arrow does not. The direction indicated by the white arrow is preferably any of the X direction or the γ direction. Here, the exposure substrate 156 is moved in the X direction. Further, the inclination angle 0 t may be appropriately set in accordance with the size of the beam spot image formed on the exposure substrate or the interval between the beam spot images. <Formation of Continuous Pattern in X Direction> Fig. 22 is a sequence diagram when a continuous pattern is formed in the X direction. 22(a) to 22(a_4)' are positional relationship diagrams of the exposure substrate 156 and the beam spot image 26A; and FIGS. 22(b1) to 22(b-4) show the movement by the exposure substrate 156. Subject to exposure. 22(a) to 22(a-4) and 22(b1) to 148200907596 Fig. 22(b-4)' also shows a part of the beam spot image 260 formed on the exposure substrate 156 and the exposure substrate 156. Part of the map. When a continuous pattern is formed in the X direction, a predetermined one of the beam spot images 260 is used. In Fig. 22, the beam spot image 260a is indicated by a black ball. That is, the light from the light source 210 is irradiated onto the exposure substrate 156 only at the position of the beam spot image 260a. As shown in Fig. 22(a1), "the exposure substrate 156 is first positioned at a desired position", and one beam spot image 260a is formed on the exposure substrate 156, so that the exposure substrate 156 is formed as shown in Fig. 22 (b1). Exposure point 262a. Next, the exposure substrate 156 is moved by a predetermined distance in the -X direction, and the beam spot image 260a is formed on the exposure substrate 156 at this position, so that one exposure point 262b is formed on the exposure substrate 156 as shown in Fig. 22 (b-2). . Further, the predetermined distance 'moved in the -X direction' may be appropriately set in accordance with the size of the dot image formed on the exposure substrate 1 56 or the interval of the dot image. In the same manner, the exposure substrate 156 is sequentially moved in the -X direction by a predetermined distance. At these positions, the beam spot image 26〇a is formed on the exposure substrate 156, so that it can be as shown in Fig. 22 (b-3) and Fig. 22 (b). As shown in FIG. 4, one exposure point 262c and 262d are sequentially formed on the exposure substrate 156. As shown in the drawing, while the exposure substrate 156 is moved by a predetermined distance in the -X direction, a beam spot image of 2 60 a is formed on the exposure substrate 1 $6, and the pattern can be continuously formed on the exposure substrate along the X direction. 156. In the above-mentioned Figs. 22(b1) to 22(b-4), in order to make the exposure points 262a to 262d formed on the exposure substrate 156 more specific, the moving distance of the exposure substrate 156 is made to be equivalent, however, As described above, the large moving distance ' of the spot image of the exposure substrate 156 149 200907596 1 56 may be set as appropriate in accordance with the interval formed on the exposure substrate or the dot image. <Formation of a continuous pattern in the γ direction> Fig. 23' is a sequence diagram when a continuous pattern is formed in the γ direction. 23U-1) to 23(a-3), the positional relationship between the exposure substrate 156 and the beam spot image 26A is shown in Fig. 23(b-3), and is exposed by the movement of the exposure substrate 156. Where. 23(6)) to 23(4) and Figs. 23_ to 23(b-3), similarly, a part of the beam spot image formed on the exposure substrate 156 and a part of the exposure substrate 156 are shown. When a continuous pattern is formed in the Y direction, two or more dot images of a plurality of beam spot images 60 are used. In Fig. 23, an example in which three beam spot images 260a to 260c are used is indicated by a black ball. That is, the light from the light source 210 is irradiated onto the exposure substrate 156 only at the point where the beam spot is 26 〇 a to 26 〇 c. As shown in Fig. 23 (a-1), first, the exposure substrate 156 is positioned at a desired position, and one beam spot image 26 〇 a is formed on the exposure substrate 156, so that exposure is as shown in Fig. 23 (b1). The substrate 156 forms i exposure points 264 and then the exposure substrate 156 is moved by a predetermined distance in the +X direction. This position is a position where the beam spot image 260b is side by side with the exposure point 264a in the +Y direction. At this position, the beam spot image 26 〇 b is formed on the exposure substrate 156, and one exposure dot 264b can be formed on the exposure substrate ι 56 as shown in Fig. 23 (b-2). As described above, the predetermined distance by which the exposure substrate 156 is moved in the +X direction enables the beam spot image 260b to be spaced apart from the exposure point 264a in the +Y direction. 150 200907596 In the same manner, the exposure substrate 156 is sequentially moved by a predetermined distance in the direction, so that the beam spot image is arranged side by side with the exposure point 264a in the +Y direction to form a beam spot image on the exposure substrate 156, thus, as shown in FIG. 23(b) As shown in -3), one exposure point 26 is formed on the exposure substrate 156 in sequence. As described above, while the exposure substrate 156 is moved by a predetermined distance in the ^ direction, two or more beam spot images 26 〇 are formed on the exposure substrate (5), and a continuous pattern can be formed on the exposure substrate 156 in the γ direction. . In the above-described FIGS. 230M) to 23(b-3), in order to make the exposure m264c formed on the exposure substrate W clearer, the moving distance of the exposure substrate 156 is made to some extent, however, as described above, the movement of the exposure substrate 156 is performed. The distance is such that the point image can be formed in a side-by-side position along the +γ direction. β In the above example, whether the continuous pattern is formed in the X direction or the continuous pattern is formed in the x direction, the light emitted from the light source 210 is irradiated on the exposure substrate 156 for a predetermined short time, only a plurality of Although i of the beam spot images, the light emitted from the light source 21 can be irradiated onto the exposure substrate 156 by forming a plurality of (two or more) beam spot images 26 既 in a predetermined short time. Thereby, the efficiency of the process can be achieved, and the time required for pattern formation can be shortened. Further, in the projection exposure apparatus 2 shown in FIG. 20, a plurality of digital micromirror elements 23A and i projection lenses 150 corresponding thereto are included, but a plurality of digital micromirror elements can also be used. And a plurality of pairs of projection lenses formed thereon, and a plurality of projection lenses are used to form a plurality of beam spot images 26 在 at a predetermined time point. By this, it is possible to further shorten the time required to form a pattern by 151 200907596. <Generation of Rank Difference> As described above, in the projection exposure apparatus 2 of the second embodiment, the same substrate mounting table 170 as that of the projection exposure apparatus 1 of the i-th embodiment is used. Therefore, similarly to the projection exposure apparatus 1A, a movement error between the X-direction moving stage 172 and the γ-direction moving stage 174 occurs. Also, Abe error will occur as well. Furthermore, the exposure substrate 156 (156a' 156b, and 15 is also the same. Therefore, the displacement error of the reference mark caused by the deformation of the exposure base 156 or the error generated when the reference mark is formed is also In the case of the projection exposure apparatus 1 相同, the same occurs. ★ This is the projection exposure of 2G0 in the second embodiment, and the various processing described in the first embodiment is, for example, the entire method of FIG. The exposure processing performed, the processing using the reference substrate (10) or (10) shown in FIG. 14 or FIG. 17, the exposure processing using the reference substrate sin or the soup shown in FIG. 18, and the alignment shown in FIG. The positioning method and the positioning process of the entire κ application push, or the displacement XY in the X direction is replaced by the difference ΔXM(n) in the X direction, and the rotation Θ is replaced by the difference ΔXM(8) in the γ direction, or as shown in FIG. In the above-described second embodiment, the first position correction mechanism is configured by the control device 99, and the above-described processing can be performed in the same manner as in the first embodiment. Second position correction mechanism, reference Plate position control mechanism, reference substrate position storage mechanism, reference substrate exposure mechanism, reference substrate position correction calculation stored in ing. w, storage mechanism, exposure substrate position control mechanism, 152 200907596 exposure substrate position library structure, and linear error correction [Embedded substrate reference mark position storage device &quot;&quot; 10,000-frame structure. [Simplified illustration of the drawings] Fig. 1 is a schematic diagram of the implementation of the present invention. Figure 2 (4) of the right-hand exposure device 1 is used for X-direction movement Detailed front view of the stage 172"4 - direction shift stage view. Detailed view of the X direction moving stage 172 is shown in Fig. 3 for moving in the Z direction, ^ ^ Detailed side view of the stage 176, U) Sun spleen The second 174 is positioned at the end of the -γ direction; _=

The direction of movement is carried from the end of the γ4 in the +γ direction. (1) Fig. 4 (a) and (b) show examples of exposure substrates [56a and (10). 5(a), (b), and (4) show the M @ ^ w sample of the reference mark *p in the first step of the step by piece method. The position detection of the position, the exposure of the substrate, and the step of exposure to the exposure substrate 156a. Position adjustment, bear; 6(4)-(b).., (C), (4) is the second step of the step by piece method, the position detection of the reference mark, the position of the exposure substrate 156b, and the exposure A step diagram of exposure of the substrate 156b. Fig. 7 (4), (8), and (c) show the positional detection of the reference mark in a slice-by-slice manner, the positional adjustment of the exposure substrate (10), and the exposure of the exposure substrate 156b. Fig. 8 is an example of an exposure substrate which is detected in a one-piece manner. FIG. 9 is a process flow 153 when a substrate 156c is exposed. FIG. 10 is a reference mark for four exposure regions (ER1, ER4, ER9, or ER14) of the exposure substrate 156c mounted on the stage 177. A state diagram when rM1 and the reference mark RM2 are photographed. Fig. 11 (a), (b), (c), (d), (e), and (f) are diagrams showing the respective states of six parameters Sx, Xy, β, ω, Ox, and Oy. Fig. 12 (a) is a schematic front view showing a reference substrate 丨 58 8a; and (b) is an enlarged view of a reference mark formed in a plurality of regions of the reference substrate 158a. 13(a) and (b) are each of the twelve regions ER1 to ER12, and one of the plurality of fiducial markers is used. Fig. 14 is a flow chart showing the position of the reference mark pair for determining the reference substrate 158a. 15(a) is an enlarged view showing a reference mark of one region of the reference substrate 158b, and FIG. 16(4) shows a reference pattern formed on the reference position reticle; (b) Not benchmark 樟$, +琛. 158b, wherein the maps ORM1 to ORM9 corresponding to the non-covered portion 44 have been shoveled like φ and -1 to P to all regions ER1-ER12 of the reference substrate 158b; (C) ^ ± ^ ^ , ) The transfer state of the pattern RM1 to 0RM9 when the center is displaced. ' and the process of transferring the reference pattern to the reference substrate _. The flow of the position of the fiducial mark pair of the earth plate 158b is corrected by the flow 15 8b to correct the movement error of the X 174. U U indicates that the reference substrate 15 is used to move to the moving stage 172 in the Y direction and in a one-chip manner. The flow body 154 200907596 The method and the weight of the whole film method are used to determine the position of the Y-direction moving stage 174. FIG. 19 shows the step-by-piece X-direction moving stage 127 and processing. Flow chart. Fig. 20 is a schematic view of the second embodiment of the present invention. The projection exposure apparatus 100 of the embodiment of Fig. 2 is formed by exposing a beam point image 260 of the "a" to the mate 1 and the exposure substrate 156. Figure. Fig. 22 (a 1) to (a_4), (7), ... (4) are sequential diagrams in which a continuous pattern is formed in the X direction. Fig. 23 (a 1) to (a_3) and (b-i) to (b_3) are sequence diagrams in which a continuous pattern is formed in the γ direction. Fig. 24 is a view showing a state in which the exposure substrate is deformed, (a) shows a case where the deformation of the exposure substrate is approximated by a quadratic formula, and (b) shows a case where the deformation is similar in the third-order equation. Fig. 25 is a processing sequence diagram of the smart whole chip mode. Fig. 26 is a schematic view showing an example of an exposure substrate used in the processing of the smart whole film method. Fig. 27 is a schematic view showing an exposure substrate, showing that the detection of a part of the exposure substrate reference mark is omitted. Fig. 28 is a schematic view showing an exposure substrate, and the detection of a part of the exposure substrate reference mark is omitted. Fig. 29 is a schematic diagram showing an example of a detection area error parameter value. Fig. 30 (a), (b), and (c) are schematic diagrams showing an example of correction of the position error when processed by the smart whole chip method. 155 200907596 Figures 31 (a) and (b) are schematic diagrams of conventional alignment optical systems. Fig. 32 (a), (b), (4), and (4) are schematic views of the mark positions of the exposure regions. Fig. 33 (a), (b), (c), and (4) are schematic views showing the arrangement of other mark positions in the exposure region. Fig. 34 (a), (8), (4), and (4) are schematic views showing positions where the curve is approximated in the exposure region. Fig. 3 5(a) (b) (c), (d) are a schematic diagram of a curve group facing each other in an exposure region, a curve group orthogonal thereto, and an exposure position coordinate. Fig. 36 (a) and (b) are schematic views showing an example of the shape of the exposure region to be corrected. [Main component symbol description] 100: Projection exposure device 11 〇: Light source 142: reticle 150: Projection lens 156 156a ' 156b ' 156e ' Wd, 156e, 156f: exposure substrate 158a, 158b: reference substrate 16 〇, 16 〇 a, 16〇b: alignment optical system 162: image processing apparatus I70: substrate loading stage 172. X-direction moving stage I74: γ-direction moving stage 156 200907596 176 : 177 : 199 : 200 : 210 : 230 : RM1 ER ·· Z-direction moving table (loading table) Control device Projection exposure device Light source digital micro-mirror element, RM2, RM3, RM4 : Exposure area, detection area reference mark 157

Claims (1)

200907596 X. Patent application: One projection exposure method is to use a projection exposure device to irradiate exposure light onto a patterned reticle to project the pattern image onto the exposure substrate loaded on the loading stage. The projection exposure apparatus comprises: a driving stage, wherein the loading stage for loading the substrate is moved and positioned at nine predetermined positions; the V/stand signal output mechanism is disposed on the driving stage, and the output indicates the loading a position signal of the position of the stage; a reference mark detecting means for detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate; and a position determining means according to the position signal The position 1 is determined to determine the position of the loading platform; and the exposure substrate has at least 9 exposures and fields for projecting the pattern image; and the substrate is exposed to at least 9 for detecting the deformation of the substrate. a detection area; and an exposure substrate reference mark ' indicates a reference position of the at least nine exposure areas, and indicates the at least a reference position of the nine detection areas; and comprising the steps of: controlling the substrate position control step, after the exposure substrate is mounted on the loading stage as the substrate, the loading stage is sequentially driven by the driving stage 158 200907596 The private movement is clamped to the four predetermined exposure substrate detection positions to V; the exposure substrate position storage step is to obtain the position of the loading platform at each of the exposure substrate detection positions from the position signal and store it; &quot;# The reference mark position calculation step is performed at each of the material light substrate detection positions, and the reference mark detecting means detects the exposure area reference mark, and calculates the exposure substrate reference mark ^ position based on the detection result and the position of the loading stage; In the error parameter value calculation step, the total area error parameter value characterized by the error of the displacement according to the exposure substrate reference mark is calculated based on the position of the exposure substrate reference mark based on the position of the exposure substrate reference mark; The error parameter value calculation step is performed in each of the detection areas, according to at least The position of the two exposure substrate reference marks is calculated, and the detection area error parameter value characterized by the error of the displacement according to the exposure substrate reference # is calculated. The detection area error parameter value linear component calculation step is performed in each of the detection areas. The detection region error parameter value is calculated by using a least square method to calculate a linear component of the detection region error parameter value; and the differential linear component calculation step is to calculate at least two steps of the detection region error parameter value for two adjacent detection regions. Difference, based on the difference, using a least squares method to calculate a differential linear component; a weighting coefficient calculation step in each of the detection regions, based on the detected region error parameter value, a linear component of the detected region error parameter value, and a linearity according to the difference Calculating the error information of the difference parameter of the component 159 200907596, and calculating the weighting coefficient according to the error information of the error parameter value of the detection region, the weighting coefficient is used to represent the error according to the detection region The linear component of the parameter value and the difference And a ratio of an error of the linear component to the magnitude of the error not based on the error; and an exposure substrate position determining step of calculating the position of the loading platform based on the position of the exposure substrate reference mark, the total area error parameter value, and the weighting coefficient The target position, the loading station is positioned at the target position. 2. The projection exposure method according to claim 1, wherein the detection region parameter is an expansion and contraction in the X direction, an expansion and contraction in the γ direction, a rotation in the x direction, a rotation in the x direction, a shear deformation in the direction, and ¥ At least one of the shear deformations of the direction. The projection exposure method of claim 2 or 2, wherein the error information for calculating the detection region error parameter value for each of the detection regions is based on the difference order of the linear component linear component of the detection region error parameter value The sum of the cumulative sums, the standard deviation of the values obtained by subtracting the difference parameter values; the sub- and the target position of the loading station mislocated from the detection area according to the difference, the linear component of the error parameter value in the detection area and the The sum of the cumulative sums of the differential linear components and the error value of the detected region are calculated based on the weighting coefficients. 4. A projection exposure apparatus for irradiating exposure light onto a patterned reticle to project the pattern image onto an exposure substrate; and comprising: driving the carriage at a predetermined position for loading the substrate The loading station is positioned at least 160 200907596 position signal output mechanism, disposed on the driving stage, and outputs a position signal indicating a position of the loading station; a reference mark detecting mechanism for detecting the exposure substrate formed on the exposure substrate a reference mark or a reference substrate reference mark formed on the reference substrate; and a position determining unit that determines a position of the loading stage based on a position indicated by the position signal; f the exposure substrate has a projection image for the image At least 9 exposure areas; 忒 exposure substrate having at least 9 detection areas for detecting deformation of the substrate; The exposure substrate reference mark indicates a reference position of the at least nine exposure regions and indicates a reference position of the at least nine detection regions; and the β-home position determining mechanism includes the following mechanism: an exposure substrate position control device #, After the exposure substrate is loaded on the loading platform as the substrate, the loading stage is sequentially moved by the driving stage to at least four predetermined exposure substrate detecting positions; the exposure substrate position storage device is (4) The signal is obtained at the position of the job table at each of the exposure substrate detection positions, and is stored; the reference mark position calculation means is configured to detect the exposure area reference mark by the reference mark detection means at each of the exposure substrate detection positions Then, based on the detection result and the position of the loading table, the position of the exposure substrate reference mark is calculated; the total area error parameter value homing and dissipating mechanism is based on the entire exposure substrate, 161 200907596 method, and the calculated area is based on Exposing the position of the substrate reference mark, using the smallest, according to the exposure substrate reference The error of the displacement is the measurement unit error parameter value calculation mechanism, and in each of the detection areas, the position of the reference mark of the exposure substrate is calculated, and the error of the displacement according to the reference mark of the exposure substrate is calculated. Detection area error parameter value; ^ Detection area error parameter value Linear component I The current tool extraction mechanism is used in each detection area. According to the error value of the detection area, the linear component of the error parameter value of the detection area is calculated by the least square method. According to the differential knife linear component calculating means, the detection zone calculates a difference of at least the order of the detection region error parameter values for the two adjacent domains, and calculates a difference linear component using the least square method; the weighting coefficient calculation mechanism And in each of the detection regions, according to the detection region error parameter value 'the linear component of the detection region error parameter value, and the error information corresponding to the differential order component of the differential linear component, and then according to the detection The error information of the parameter value is used to calculate a weighting coefficient, and the weighting coefficient is used According to the linear component of the detection region error parameter value and the error of the differential linear component, and the magnitude of the error not based on the error; and the exposure substrate position determining mechanism, according to the position of the exposure substrate reference mark, the overall error The parameter value and the weighting coefficient are used to calculate a private position of the loading station, and the loading station is positioned at the target position. 5. The projection exposure apparatus of claim 4, wherein the inspection 162 200907596 measures the regional parameters, your γ ^ , the expansion and contraction of the direction, the expansion and contraction of the γ direction, the rotation of the χ direction, and the rotation of the Υ direction v ^ At least one of twirling, shear deformation in the X direction, and shear deformation in the γ direction. 6. If the projection exposure apparatus of the fourth or fifth patent application scope is applied, ',,,,,,, B.3⁄4 f | Μ, the error information of the error value of the detection area is calculated for each 5 hai detection area, The linear component of the I-domain error parameter value and the sum of the difference JJ of the linear component of the differential tool and the white number of the left knife P are subtracted from the error value of the detection region. The standard deviation of the value; the target position of the unit is the linear component of the error parameter value of the detection region and the sum of the sum of the difference linear components, and the value of the error parameter of the region, according to the weighting Calculated by the coefficient. 7. An alignment method for detecting a mark position on a printed circuit board when a pattern on a reticle is projected and exposed to a rectangular printed circuit board through a projection optical system, and determining an exposure position based on the mark position information, wherein The method includes the following steps: dividing a plurality of mark information arranged in parallel with one side of the rectangular substrate into at least two groups corresponding to one pair of opposite sides. According to design coordinates information of each mark in the group, and The detection information of each mark position determines the number of equations having a straight line or the curve information obtained by the approximation; 'For each of the two types determined by the two groups, substituting the design coordinate value for exposure to calculate 2 Position information; and the exposure position of the two sets of mark coordinates in the axial direction orthogonal to the calculated coordinate axis of the exposure coordinate is used to determine the exposure position. 163 200907596 8. The alignment method, when the pattern on the reticle is projected and exposed to the rectangular printed circuit board through the projection optical system, the mark position on the printed circuit board is detected and the exposure position is determined according to the mark position information. The method includes the following steps: dividing a plurality of pieces of mark information arranged in parallel with one side of the rectangular substrate into at least two groups corresponding to one pair of opposite sides; according to the design of each mark in the group Coordinate information, and the detection of each marked position, to determine the number of equations with a straight line or the approximated curve information; ^ ' group listening to each of the number. '丨 ~ eight goods encounter exposure area The design coordinate value of the diagonal winding ^^, 〇 to the parent 2钿, the magnification of the exposure area and the rotation information are calculated according to the 苴 苴 八 and the use of the exposure zen $ $ @ 铋 与 与 正交The direction of the 9 groups of markers is relative to the inner division of the 2 points of the knife to determine the exposure position. 9. If you apply for the alignment method of the 7th or 8th item of the patented gluten garden, the 配置 is arranged in a plurality of marks with the 2 thin τ ^ which will be in the direction of the flat, , v, and parent directions. Divided into less than 1 oblique aunt + m τ &lt;lean news, to the corresponding treatment of the side of each other. And enter the same 〗 〖. For example, the line or use determined in the alignment group of the 7th or 8th patent application scope, wherein the information of the curve determined by the number of the mark is in the case of the case 2 2 = Department - sub-type, marked as. In the case of 4 or more points, it is 3 times. 11 · For example, the alignment method of the 8th item of Shenjing Patent Range No. 2 is included in 164 200907596 仃 before the alignment is determined by (4) the number of points to be calculated: a step up to the formula; the step 'selects the mean value of the filaments corresponding to the sides opposite to each other as determined by the straight line approximation "2 times or a straight line like the first order, or 2 times the approximation of the 2nd curve formula. A aligning method according to claim 7 or 8, which comprises the step of detecting the mark of three or more points in each of the groups of the processing batch of the printed circuit board; The value of the approximation coefficient of the second and third times in the approximation or curve approximation reduces the number of points in the second and subsequent bars compared to the number of points in the slice. 13. An exposure apparatus for projecting and exposing a pattern on a reticle through a projection optical system to a rectangular printed circuit board, comprising: a drive stage for loading a loading platform for loading the printed circuit board Positioning and outputting; a position signal output mechanism disposed on the driving stage and outputting a position signal indicating a position of the loading stage; a mark position detecting mechanism for detecting a mark position formed on the printed circuit board; and aligning The control mechanism determines the exposure position according to the detection information of the mark position obtained by the mark position detecting mechanism, and controls the driving stage according to the determined exposure position and the position signal; the alignment control mechanism comprises: The information classification mechanism is configured to divide at least a plurality of mark information arranged in parallel with the i side of the rectangular substrate into at least two groups corresponding to the side 165 200907596 opposite to each other; the number determination mechanism 'based on the group The design coordinates of each mark and the detection information of each mark position 'to determine whether there is a straight line or profit The number of the curve information obtained by approximating the enthalpy; the position of the position calculation mechanism 'for each of the two types determined by the two groups, substituting the design coordinates of the exposure to calculate two position information; and exposure The position determining mechanism is a relative internal information of the two sets of mark coordinates of the exposure coordinate and the calculated direction of the axis of the coordinate axis, and the exposure position. Water exposure device 4. An exposure device is a projection exposure device that projects a pattern on a reticle through a projection optical system to a rectangular printed circuit board, and is characterized in that the drive stage is moved and positioned; a position signal output mechanism of the loading platform of the printed circuit board, a position signal disposed on the driving stage and at a position of the loading station; a position detecting mechanism is not marked for detecting a mark position formed on the printed circuit; and The quasi-control mechanism determines the exposure position based on the detection information of the four positions obtained by the mark position detecting mechanism, and controls the driving stage according to the determined H exposure position and the position signal; the alignment control mechanism includes : Marking the information classification mechanism, a plurality of mark information will be arranged in parallel with the i side of the rectangular substrate, at least divided into two pairs of opposite sides of each other; 166 200907596 two groups; the number determining mechanism, according to each of the groups Mark the design coordinate information, and the detection information of each mark position to determine the straight line or use the approximation The method of calculating the curve information; the magnification and rotation information calculation means, for each of the two types determined by the two groups, substituting the design coordinates of at least two ends of the exposure direction of the exposure region for exposure, by the difference Calculate the magnification and information of the (4) light area; and the exposure position determining mechanism determines the exposure position by using the internal coordinate information of the exposure coordinates and the two sets of coordinate coordinates of the calculated axis in the positive axis direction. The exposure apparatus of claim 13 or 14, wherein the plurality of marker information arranged in parallel to the side orthogonal to the two groups is divided into at least two groups corresponding to i facing each other. And the same treatment is applied. In the exposure apparatus of claim 13 or 14, wherein the straight line determined in the group or the curve information obtained by the approximation is determined by the number of the mark, the mark In the case of 2 o'clock, the system is a sub-type, a standard, and the case of 3 o'clock is a second-order type, and when it is marked as 4 or more points: the person's sub-form. The package is to be calculated or straight. For example, in the exposure apparatus of claim n or 14 of the patent range, before the alignment is performed, the step up to the order is determined in a manner irrelevant to the number of points of the mark; the step is to select a straight line approximation with the pair of i The average value of the corresponding equations on the opposite side of the jade 167 200907596 line approximation of the first-order formula, or the second-order curve approximation of the second-order formula. 18. An exposure apparatus according to claim 13 or 14, comprising a step of processing a batch of printed circuit boards, and performing a mark detection of more than three points for each of the groups; the step of performing a linear approximation or When the curve is approximated, the values of the approximation coefficients of the second and third times are such that the number of mark detection points after the second slice is smaller than the number of dots of the first slice. 19) A projection exposure apparatus that irradiates exposure light onto a patterned reticle to project the pattern image on an exposure substrate, comprising: a drive stage, a loading stage for loading the substrate Positioning at at least four predetermined positions; a position signal output mechanism disposed on the drive stage and outputting a position signal indicating a position of the loading stage; a reference mark detecting mechanism for detecting an exposure substrate reference formed on the exposure substrate a mark or a reference substrate reference mark formed on the reference substrate that can be exposed; and a position determining mechanism that determines a position of the loading table based on a position indicated by the position signal; the exposure substrate reference mark is used The reference substrate reference mark is composed of at least four marks indicating a reference position of the reference substrate; the position determining means includes: the first position branch a positive mechanism for correcting the loading station produced by the 168 200907596 when moving the loading platform a position error; and a second position correcting mechanism for correcting a linear component error in a position error of the exposure substrate reference mark; the first position correcting mechanism comprising: a reference substrate position control mechanism, the reference substrate being used as the substrate After being loaded on the loading platform, the loading stage is sequentially moved by the driving stage to at least three predetermined reference substrate detecting positions; and the reference substrate position storing mechanism is obtained from the position signal. The position of the loading table at the substrate detection position is stored; the reference substrate exposure mechanism is used for each of the reference substrate detection positions, and the light is used as a reference reticle to define the reference mark formed on the reference reticle. a substrate on which the image of the line reference mark is formed on the reference substrate; and a reference substrate position correction calculation storage means for detecting the mark 2 reference substrate reference mark formed on the reference substrate by the reference mechanism According to the detection result, the reticle = "the image and the relative position of the 忒 reference substrate reference mark are calculated... The position of the loading platform is calculated and stored in the loading position correction data for correcting the error. The second position correcting mechanism includes: an exposure substrate position control mechanism name mounted on the mounting substrate to be broken as the substrate, After the carrier port, the hunting platform is positioned by the driving stage, and the loading device is positioned at the exposure substrate detection position determined by the loading station position correction data; 169 200907596 exposure substrate position storage mechanism, Obtaining, from the position signal, a position of the loading table at each of the exposure substrate detection positions and an "exposure-based sub-marking mark position storage mechanism" for detecting the exposure substrate by the reference mark detecting mechanism at each of the exposed substrate release positions The reference mark 'then calculates the position of the exposure substrate reference mark according to the detection result and the position of the loading table and stores it; and 'in the error correction nose transport mechanism, according to the storage mechanism at the exposure substrate reference mark position storage mechanism The position of the exposure substrate reference mark is calculated using a least squares method to correct the linearity Linear error correction for compositional errors. 20: a projection exposure apparatus for exposing exposure light to a patterned line; 1 'projecting the pattern image on an exposure substrate, comprising: a drive stage for loading a substrate The loading station is positioned in at least four predetermined positions; a position signal output mechanism is disposed on the driving stage, and outputs a position signal indicating a position of the loading station; and a reference mark detecting mechanism for detecting the formed on the exposure substrate Exposing the substrate reference mark or the reference substrate reference mark formed on the reference substrate, and the position determining means for determining the position of the seismic stage based on the position indicated by the position signal; It is composed of at least four marks for indicating the reference position of the exposed substrate; 170 200907596 »Hai base substrate reference mark, at least 'quasi-position', used to indicate the base of the reference substrate; The position determining mechanism includes: a first position correcting mechanism, which is produced when the wearing father is moving the dressing table The positional error of the stage; and the second position correcting mechanism, the first position correcting mechanism for exposing the substrate reference mark by the linear component error in the position error includes: a reference substrate position control mechanism, and the crack is carried on the display After the stage, the M = reference substrate is moved as the substrate to be positioned at at least three predetermined dynamic loads; the loading platform sequentially picks up the dry substrate detecting position; the reference substrate position storage mechanism, and the 基板# a L system The position signal is obtained at the detection position of the 夂哕 reference substrate. The position of the sea battle port is stored and stored; the reference mark position of the ground reference plate stores the detection position, and the reference substrate base is detected by the U-substrate (4) 4 μ detection mechanism, and then Detecting the position of the buckling reference substrate reference f 2 and (4) the stage, calculating the position of the 丞 己 且 and storing it; and calculating the position of the reference substrate.   ^, the storage mechanism, according to the reference base; the position of the quasi-marking + the substrate base only stops the position misalignment position correction data of the loading station and stores it; loading the second position correction mechanism comprises: exposing the substrate position control mechanism After being loaded on the loading platform, the substrate is moved as the substrate to be positioned at least Π 171 200907596 according to the position correction data of the loading table; the position of the exposed substrate is detected; 'Acquiring and storing the position of the loading table at each of the exposure substrate detection positions from the &amp;signal; the exposure substrate reference mark position storage mechanism is at each of the exposure substrate detection positions by the reference mark detecting mechanism Detecting the exposure substrate reference mark, and calculating the position of the exposure substrate reference mark according to the detection result and the position of the loading stage, and storing the error; and recording error correction operation mechanism, storing according to the position of the exposure substrate reference mark stored The position of the exposure substrate reference mark of the mechanism is calculated using the least square method To correct the linear error correction of the linear component error. 21. A projection exposure apparatus comprising: a digital micromirror element having a plurality of mirrors, and wherein a direction of reflection of light incident on the plurality of mirrors is dependent on the plurality of mirrors; and a microarray lens And a plurality of microlenses corresponding to the plurality of mirrors respectively; and the point image formed by the microarray lens is projected onto the exposure substrate, wherein the method comprises: driving the stage, which is used for loading the substrate The loading station is positioned in at least three predetermined positions; the position signal output mechanism is disposed on the driving stage, and outputs a position signal indicating a position of the loading station; and a reference mark detecting mechanism for detecting the formed on the exposed substrate Exposing the substrate reference mark or the reference substrate reference mark formed on the reference substrate that can be exposed; and 172 200907596 position determining mechanism, determining the position of the loading table; and determining the exposure substrate reference mark based on the position indicated by the position signal a reference substrate base composed of at least three marks indicating the position of the exposed substrate The mark is formed by at least three marks indicating the position of the reference substrate. The position determining mechanism includes: a first position correcting mechanism, wherein the position error of the loading table is used to correct the production when the loading table is moved a second position correction mechanism for correcting a linear component error in a position error of the exposure substrate reference mark; the first position correction mechanism includes a reference substrate position control mechanism, and the reference substrate is mounted as the substrate After the loading platform, the loading table is sequentially positioned to move at least three predetermined reference substrate detecting positions by the driving stage, and the reference substrate position storage mechanism is obtained from the position signal and detected on each of the reference substrates. The position of the loading table is stored and stored; the reference substrate exposure mechanism is for each of the reference substrate detection positions, and the exposure light is irradiated onto the reference reticle to form a reticle reference formed on the reference reticle Marking a projection on the reference substrate to form an image of the reference mark on the reference substrate; and a reference substrate position The positive operation storage unit detects the image of the reticle &amp; mark and the reference mark formed on the reference substrate by the reference mark detecting means, and calculates the reticle reference 173 200907596 according to the detection result The relative position of the reference substrate and the position of the loading table are "calculated for the purpose of the calculation", and then the position of the loading table is corrected according to the error (4) and is given to the fourth parent; the second position of the position of the loading table The calibration mechanism includes: the exposure substrate is controlled to be spread, and the exposure substrate is used as the substrate by the drive stage, and the loading stage is sequentially moved and clamped to the exposure substrate according to the loading stage. The at least three light-receiving position storage mechanisms are obtained from the position signal to obtain the position of the loading table at each of the substrate detecting positions and stored; the exposure substrate reference mark position # 、, 彳 储存 储存The mechanism detects the exposed substrate reference "self" by using the reference mark detecting mechanism at each of the exposed substrate=positions, and then according to the detection result and the loading The position of the exposure substrate reference mark is calculated and stored; and the "linear error correction calculation mechanism" uses the least square position based on the position of the exposure substrate reference mark stored in the exposure substrate reference The method calculates a linear error correction data for correcting the error of the linear component. 0, 22· a projection exposure apparatus, comprising: a digital micromirror element, having a plurality of mirrors, and capable of emitting light to the plurality of mirrors The reflection direction is determined by the plurality of mirrors respectively; and the microarray lens has a plurality of microlenses corresponding to the plurality of mirrors respectively; and the point image formed by the microarray lens is projected onto the exposure substrate, The invention comprises the following features: 174 200907596 The drive stage 'positions the loading stage for loading the substrate in at least three predetermined positions; the position signal output mechanism is disposed on the driving stage, and outputs the position indicating the position of the loading stage a signal mark detection mechanism for detecting exposure formed on the exposure substrate a substrate reference mark or a reference substrate base formed on the reference substrate; and a π position determining means for determining a position of the loading stage based on a position indicated by the position signal; and the exposure substrate reference mark is used to indicate the Forming at least three marks of a reference position of the exposure substrate; the reference substrate reference mark is composed of at least three marks indicating a reference position of the reference substrate; and the position determining means includes: a first position correcting mechanism; a second position correction mechanism for correcting a linear component error in a position error of the exposure substrate reference mark; and a second position correction mechanism for correcting a position error of the loading table generated when the loading table is moved; The reference substrate position control mechanism is mounted on the loading table as the substrate, and the loading stage is sequentially positioned to move at least three reference substrates. Detection location; The reference substrate position storage mechanism acquires and stores the position of the loading table at each of the reference substrate detection positions from the position signal; 175 200907596 The reference substrate reference mark position storage mechanism is at each of the reference substrate detection positions by The reference mark detecting means detects the reference mark of the reference substrate, and based on the detection result, the position of the reference mark of the reference substrate is calculated and stored, and the reference substrate position correction calculation storage means Calculating and storing the loading table position correction data for correcting the position error of the I stage based on the position of the reference substrate reference mark; the second position correcting mechanism includes: an exposure substrate position control mechanism, which is exposed at the pastoral edge After the substrate is mounted on the loading platform as the substrate, the 藓 访 访 amp amp amp 精 3⁄4 moving stage, the loading stage is sequentially positioned to at least 3 predetermined exposure substrate detection positions determined according to the position of the loading station and the inspection of the Gongga; the exposure substrate position is stored; The position signal is obtained at the position of the exposure substrate of each of the exposure substrates, and is stored and stored; the exposure substrate reference mark position is stored in the detection mechanism, and the exposure substrate detection position is By using the reference 椤 仏 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , And the linear error correction of the real machine medium, the dissimilar structure is based on the position of the substrate on which the substrate is stored, and the position of the reference mark of the substrate is also suitable for the night. The linear error correction data used to correct the linear component error is calculated using the least squares method. A projection exposure apparatus according to any one of items 19 to 22, which comprises a projection light station m for illuminating the exposure light to the 176 200907596 exposure The reference mark detecting mechanism 'includes an alignment optical system disposed between the projection optical system and the loading stage; 忒 aligning the optical system, and irradiates the non-exposure light to the exposure substrate reference e or the reference substrate reference mark Detecting the exposed substrate reference mark or the reference substrate reference mark; the 忒 alignment optical system is positioned at the detection position when detecting the exposure substrate reference mark or the reference substrate reference mark; at the end of the exposure substrate reference mark Or after the detection of the reference substrate reference mark, it is positioned at a retreat position that is retracted from the detection position. ^ 24. The projection exposure apparatus according to any one of claims 19 to 22, wherein the loading position correction data is calculated based on an average value of positions of the exposure substrate reference marks at a plurality of reference substrate detection positions. . ^ 25. The projection exposure liquid according to any one of the items of the invention of the present invention, wherein the reference substrate reference mark is formed at a grid-like intersection in the reference substrate at a predetermined interval; ▲ » Hai loading station The position correction data is calculated based on the position of the reference substrate reference mark located at the grid-like intersection. The secondary and exposure substrate position control means corrects the material by using the stage position and calculates by curve approximation or interpolation. The target position at which the loading station is to be positioned positions the loading station at the target position. 26. A projection exposure apparatus according to the ninth or second aspect of the invention, comprising: 177 200907596 a projection optical system capable of changing a projection magnification of the pattern image and projecting the pattern to an edge exposure substrate; and a projection magnification determining mechanism, The projection magnification is determined based on the linear error correction data that is different from the linear error correction computing means. A projection exposure apparatus for irradiating exposure light onto a patterned reticle to project the pattern image on an exposure substrate, comprising: a driving stage for loading the substrate The loading station is positioned in at least four predetermined positions; a position signal output mechanism is disposed on the driving stage, and outputs a position signal indicating a position of the loading station; and a ground-based detection mechanism for detecting the formation Exposing the substrate substrate reference &amp; δ£, or the reference substrate reference mark formed on the reference substrate, and the position determining mechanism, determining the position of the loading table based on the position indicated by the position signal; a substrate having at least four exposure regions for projecting the image of the pattern; 曝光 the exposure substrate reference mark is an exposure region reference mark for indicating a reference position of the at least four exposure regions; a position correction mechanism for correcting a position error of the exposure substrate fiducial mark; the exposure position correction mechanism comprising: The light substrate position control mechanism is mounted on the substrate 178 200907596 on the substrate 178 200907596, and the mop moxibustion is used to sequentially position the loading table in at least four predetermined positions by the driving stage. Exposing the substrate detection position; and exposing the substrate position storage mechanism to obtain the position of the loading table at each of the exposure substrate detection positions from the position signal and storing the ITO exposure area reference mark position library, each exposure The substrate*J position is used to detect the exposure area reference by the hai reference mark detecting mechanism, and then the position of the exposure area reference mark is calculated and stored according to the detection result and the position of the loading stage; the position error processing mechanism is Storing the position of the exposure area reference mark in the exposure area reference mark position storage mechanism, and calculating the error information of the nonlinear component in the position error of the exposure substrate reference mark using the minimum flat branch; the differential error processing mechanism is stored according to the exposure The exposure area reference mark of the area reference mark position storage mechanism The difference in position, using the least square method to calculate the error information of the nonlinear component in the difference error of the position of the exposure substrate reference mark; and the exposure substrate position determining mechanism 'based on at least the error information of the two nonlinear components Weighting to calculate the target location at which the loading station is to be positioned, and then positioning the loading station at the target location. A projection exposure apparatus for irradiating exposure light onto a patterned reticle to project the pattern image on an exposure substrate, comprising: a driving stage, and a loading platform for loading the substrate Positioning in three predetermined positions; 179 200907596 position signal output mechanism, disposed on the driving stage, and outputting a position signal indicating the position of the loading stage; a reference mark detecting mechanism for detecting the exposure substrate formed on the exposure substrate a reference mark; and a 'position determining mechanism' determines a position of the loading stage according to a position indicated by the position signal; the exposure substrate has at least three exposure areas for projecting the pattern image; and an exposure area is used to correct the The exposure substrate reference mark is an exposure area reference mark indicating a reference position of the at least one field; The position-receiving mechanism of the home position includes an exposure position correction mechanism, and the position error of the exposure substrate reference mark; the war exposure position correction mechanism includes: · exposure substrate position control.   _, after the material light substrate is mounted on the loading platform as the substrate, the burial earth plate is driven by the driving stage, and the loading table is positioned to move at least 3 彳 g g jf6 The exposure substrate detecting position; the exposure substrate position storage mechanism is to take the position of the mounting table of the exposure substrate detection position from the position signal and store it. ^ Exposure area reference mark position storage, detection position, by the reference rod, 糸 in each of the exposure substrate blocks ^ 己 己 detection mechanism to detect the exposure area reference ^ and then according to the detection result and the position of the 褒 stage Jin, the location of the reference mark of the lack of light area and stored.  The nose area is exposed to the linear error correction computer to mark the position of the storage area of the exposure area base == area reference thousand knows the position of the 'use the most 180 200907596 i flat method is used to correct the linear component error material; &amp; The decision-making technique is used to calculate the position of the reference mark stored in the exposure field of the exposure field.   And the knives, with the overlap control mechanism, are obtained by using the least square method, and the 5 _ telescopic and orthogonal types are replaced by the difference, and the exposure is also applied to the reticle to Moon secret first. The overlapping target position of the 3⁄4 projection image and the exposure substrate. "2:---the projection exposure method" uses a projection exposure device to illuminate the exposed substrate formed on the loading table with light; the projection exposure device includes: a driving stage, The loading platform for loading the substrate is positioned and positioned in at least three predetermined positions; the position signal output mechanism is disposed on the driving stage, and outputs a position signal indicating a position of the loading station; &quot; a reference mark detecting mechanism for detecting a reference substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate that can be exposed; and a position determining mechanism that determines the position of the loading stage based on the position indicated by the position signal The exposure substrate reference mark 'is composed of at least three marks indicating the position of the base 181 200907596 of the exposed substrate; the reference substrate reference mark is used by the _ ^ field to indicate whether At least three marks of the reference position of the reference substrate; and the following steps: Reference substrate position control step After the base substrate is mounted on the substrate as the substrate, the loading stage is sequentially positioned to move at least three predetermined reference substrate detection positions by the driving stage; the reference substrate In the position storage step, Yu Qiu obtains the position of the loading table at each of the reference substrate detection positions from the position signal and stores it. The reference substrate exposure step is performed at each of the reference substrate detection positions, and the exposure light is irradiated. The reference mark H , , ^ L ... is such that the mark formed on the reference reticle, the 'spring slice reference mark is projected on the steam, and the plate is folded to form the line reference mark on the reference substrate Like; the reference substrate position correction 瞀 左 左 左 左 左 左 左 , , , , , 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左 左The image of the slice reference mark ^ ^ 'Based on the detection result, the relative position of the reticle reference center 5 and the reference substrate is calculated, the relative position and the position of the loading table... 卞f..., the rear root (4) Set #夕_Changing out the loading station position correction data for correcting the positional difference of the loading station and storing it; the exposure substrate position control step is performed after the exposure substrate is loaded on the loading table , Su “ 》 》 亥 亥 亥 — & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & & 6 6 At least three innocent exposure substrate detection positions and exposure substrate positions are stored in a sub-step, and the position of the loading station at each of the 182 200907596 exposure substrate detection positions is obtained from the position signal and stored; First, the substrate reference mark position storage detection position 蕤, 蕤 复 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , And storing; and ... the exposure line difference &amp; positive operation step, according to the position of the exposure substrate reference mark stored in the exposure substrate reference mark: storage step The least squares method is used to calculate the linear material to correct the linear component error. Bay 30. a projection exposure method for projecting an exposure light ',, and a pattern onto a reticle formed with a pattern to project the pattern image onto the exposure substrate mounted on a loading stage; the projection exposure apparatus The device includes: a driving load σ ϋ word for loading the substrate, the mounting table is moved and positioned in at least three predetermined positions; a position signal output mechanism is disposed on the driving stage, and outputs a position signal indicating a position of the loading station; a mark detecting means for detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate; and a position determining means for determining the loading stage based on the position indicated by the position signal Position; the feature is: &quot;Heil exposure substrate reference mark, which is composed of at least three marks indicating the position of the base 183 200907596 of the exposed substrate; the reference substrate reference mark is used to indicate the reference substrate At least 3 marks of the position of the beautiful soap; the base of the soil plate and the following steps·· The position control step is, after being loaded on the loading platform, the position of the loading platform of the reference substrate detecting position by breaking the reference loading position of the reference loading substrate in the reference loading stage. ^The signal acquisition step is performed at each of the reference substrate reference mark positions: the detection position, the H &amp; ', after each of the reference substrate alignments, the iron broadcast iP i &amp; mechanism to detect the reference substrate base... According to the detection result and the position of the load A of the load A, the reference mark of the reference mark is stored and stored, and the reference substrate position correction operation is stored and stored in accordance with the reference substrate base position. Correcting the data and storing it; the m-loading substrate position control step, after being loaded on the loading platform, is used as the substrate to print &amp; a moving stage, the mounting (4) is positioned at at least three exposed substrate detection positions determined according to the loading position correction data; = the optical substrate position storing step is obtained from the position signals at each of the exposures The position of the loading table of the substrate detection position is stored; = the pure reference mark position storing step is performed by detecting the position of the exposure substrate reference 184 200907596 by the reference mark detecting mechanism on each of the exposure substrates, and calculating the exposure; And the mark, and then according to the detection result and the position of the light-receiving substrate reference mark and the storage position is stored in the cow, $ # + * 隹垓 exposure substrate reference mark Γ Γ the position of the exposure substrate reference mark, using: The linear error correction used to correct the linear component error is to use the projection exposure device, which will be reflected to the exposure base loaded on the loading platform.  A projection image forming method of a micro-array lens; the projection exposure apparatus comprises: the loading and loading station 'moving and positioning on the driving stage, which is used for loading the substrate with three predetermined positions; ▲ position signal output The mechanism is disposed on the driving stage and outputs a position signal indicating a position of the loading stage; the reference mark detecting mechanism is configured to detect an exposure substrate reference mark formed on the exposure substrate or formed on the reference substrate for exposure a reference substrate reference mark; a position determining mechanism that determines a position of the loading stage based on a position indicated by the position signal; the digital micromirror element 'having a plurality of mirrors and capable of emitting light to the plurality of mirrors The reflection direction is determined by the plurality of mirrors respectively; and the microarray lens has a plurality of microlenses corresponding to the plurality of mirrors respectively; 185 200907596 is characterized in that: the exposure substrate fiducial mark is used to indicate At least three marks of the reference position of the exposure substrate; the reference substrate reference The method is composed of at least three marks for indicating a reference position of the reference substrate, and includes the following steps: a step of using the reference substrate as the substrate, and the driving stage a predetermined reference substrate detection position; the reference substrate is positionally controlled and loaded on the loading stage, and is moved and positioned in at least three reference substrate position storage steps, and the loading platform of the reference substrate detection position is obtained from the position signal Position and storage; signing the slab exposure step 丞 substrate position correction calculation stored in the measuring mechanism, the detection is formed at the step, the substrate reference mark is detected by the reference mark, and the second::: relative position is calculated according to the detection result: := ί:= The relative position of the mark, according to the error of the load, the load position of the loader and the position of the load of the 5 loader of the 5th load table is stored and stored; the exposure substrate position control step is loaded on the The loading table H is made up of (4) "the substrate is slightly used as the substrate. Xuan* will cut the meeting, |»^ j.  j Move positioning at least 3 186 200907596 according to the loading station, I shy station in the order position of the poor material, the predetermined exposure substrate detection position; exposure substrate position storage step, 'system position signal acquisition for a long time梦 Exposure substrate detection position of the Meng Gong A &amp; now take the position of each 5 Heitz loading station and store it; Exposure substrate reference mark position storage direct storage step, in each of the exposure = position mark detection The mechanism detects the exposure substrate i, and then according to the detection result 4, «-, c ^ ^ ^ the position of the clothing battleport' is different from the position of the substrate reference and is stored, and linear The error correction operation step is inversely different, according to the storage method stored in the exposure substrate reference position storage step, 廿 &lt; 4 The position of the substrate reference mark is measured by the least square method to correct the green data. ^, friendliness becomes the error linear error correction I A projection exposure method, using a projection exposure device, the point image formed by the column (10) is projected onto the exposure substrate; ^ The projection exposure device includes: The stage, which is used to intercept the platform of the support storm, is positioned in at least three predetermined positions; ▲ a position signal output mechanism is disposed on the drive stage, and the position indicating the position of the loading stage is output a reference mark detecting mechanism for detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate, and a position determining mechanism according to the position of the &amp; signal (4) Position of the load platform; 187 200907596 Digital micromirror element having a plurality of mirrors, and the direction of reflection of light incident on the plurality of mirrors is determined by the plurality of mirrors; An array lens having a plurality of microlenses corresponding to the plurality of mirrors respectively; wherein: the exposure substrate reference The reference substrate reference mark is composed of at least three marks indicating a reference position of the reference substrate; and the following steps are included; : a reference substrate position control step of positioning the loading stage on at least three predetermined reference substrate detection positions by the drive stage after the reference substrate is mounted on the loading stage as the base stage; The reference substrate position storing step is performed, and the position of the selected A ^ loading station at each of the writing positions of the reference substrate is obtained from the position signal and stored; the reference substrate reference mark is raised to six times. The storing step is performed at each of the reference base phases, and the detection mechanism detects the reference substrate reference '圮 by the base, and then detects the Λ m ^ ^ ^ ^位置 The position of the stage, calculate the position of the reference mark of the base plate Μ to store; the reference substrate position correction h h mm # 异 储存 storage step, according to The quasi-substrate base aw , , τ is also the position error of the 3 haistation stage position and the data is stored; the exposure substrate position control step, the 5 hai exposure substrate is loaded as the base handle 188 200907596 After the loading stage, the mobile positioning position is at least three positions determined by the Shai driving stage according to the predetermined exposure substrate detection position of the loading table, and the exposure substrate position is inaccurate; The step of obtaining the position of the seismic stage at each of the exposure substrate detection positions from the position signal and storing the same; the exposure substrate reference mark position storage step is performed by detecting the reference mark at each of the exposure substrate detection positions The mechanism detects the exposure substrate reference mark, and then calculates and stores the position of the exposure substrate reference mark according to the detection result and the position of the loading stage; and the linear block production correction operation step, according to the position of the exposure substrate reference mark The position of the exposure substrate reference mark stored in the step is calculated using the least square method to correct the line Linear error correction data of the component error of the invention, wherein the projection exposure device includes a projection for irradiating the exposure light to the exposure substrate, wherein the projection exposure device of any one of claims 29 to 32 An optical system; the reference mark detecting mechanism includes an alignment optical system disposed between the projection optical system and the loading stage, and irradiates non-exposure light to the exposure substrate reference mark or the reference substrate reference mark to detect the exposure substrate a reference mark or the reference substrate reference mark; when the reference mark reference mark or the exposed substrate reference mark is detected by the reference mark detecting means, the method includes the following steps: when detecting the exposed substrate reference mark or the reference substrate reference mark 'Locating the alignment optical system at the detection position; and 189 200907596 after positioning the exposure substrate reference mark or the reference substrate reference mark to 'position the alignment optical system at a retreat position retreating from the detection position . The projection exposure method according to any one of claims 29 to 32, wherein the loading position correction data is calculated based on an average value of positions of the exposure substrate reference marks at a plurality of reference substrate detection positions. . The projection exposure method according to any one of claims 29 to 32, wherein the reference substrate reference mark is formed at a grid-like intersection in the reference substrate at a predetermined interval; 忒 loading table position correction data Calculated based on the position of the reference substrate reference mark located at the lattice intersection; and comprising the steps of: using the load station position correction data and using curve approximation or interpolation to calculate the target position of the iU to position the loading table And then position the loading station at the target location. A person who applies the projection exposure method of claim 29 or 32, and the package 3 projection optical system can change the projection magnification of the pattern image to project the case to the exposure substrate; and the following steps are included: The linear error corrected poor material calculated by the linear error correction operation step determines the projection magnification. The 37' projection exposure method uses a projection exposure device to illuminate the exposure center light onto the patterned reticle to project the pattern image onto the exposure substrate mounted on the loading station at 190 200907596; The device comprises: a driving stage, wherein the loading platform for loading the substrate is moved and positioned in at least four predetermined positions; a position signal output mechanism is disposed on the driving stage, and outputs a position signal indicating a position of the loading platform; a mark detecting means for detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate; and a position determining means for determining the position of the loading stage based on the position indicated by the position signal The exposure substrate has at least four exposure areas for projecting the pattern image, and an exposure area reference mark for indicating a reference position of the at least four exposure areas; and the following steps are included; : The exposure substrate position control step 'in the exposure substrate is used as the substrate Loading: on the loading platform|, by the driving stage, the loading table is sequentially positioned to be at least 4 彳^β 4 曝光 exposure substrate detection positions; 讦芡π w攸 into a call signal obtained in Each of the exposed substrates detects the position of the load station t of the summer 4 and stores it; the exposure area base dimension; the m detection position, and the storage step, the detection area of each of the exposure substrates is used to detect the exposure area reference 191 200907596 mark, and then according to the detection result and the position of the loading station, the position of the light area reference mark is stored. Λ + position error processing step 'according to the exposure line stored in the exposure area reference storage step The position of the fiducial mark, using the 'method to calculate the non-linear:: error information in the position error of the exposure substrate fiducial mark; the differential error processing step, according to the exposure area fiducial mark stored in the exposure area reference mark position material step The difference of the position, using the least square method to calculate the nonlinearity of the difference error of the position of the reference mark of the exposure area Error information; and an exposure substrate position determining step of applying a weighting 'at at least one of the two nonlinear components to calculate a target position to position the loading station' and then positioning the loading station at the target position . 38. A projection exposure method, using a projection exposure device, irradiating exposure light onto a patterned reticle to project the pattern image onto the exposure substrate loaded on the loading platform; The drive includes a drive stage for positioning the loading stage for loading the substrate at at least three predetermined positions; a position signal output mechanism disposed on the drive stage and outputting a position signal indicating a position of the loading stage; a reference mark detecting means for detecting an exposure substrate reference mark formed on the exposure substrate or a reference substrate reference mark formed on the reference substrate; and 192 200907596 position determining means for determining the loading based on the position indicated by the position signal The position of the stage is characterized in that: the exposure substrate has at least a field for projecting the image of the pattern; the three exposure areas of the exposure substrate reference mark are used to indicate the exposure area reference mark of the reference position of the at least field; The exposure area includes the following steps: an exposure substrate position control step in which the exposure substrate is taken After the substrate is mounted on the loading stage, the loading stage is sequentially privately clamped to at least three predetermined exposure substrate detecting positions by the driving stage; and the exposure substrate position storing step is performed from the position signal Acquiring and storing the position of the loading stage at each of the exposed substrate detection positions; and exposing the exposure area reference mark position to detect the exposure area reference by using the 6-foot reference mark detecting mechanism at each of the exposed substrates: And then 'according to the detection result and the position of the positional light reference mark of the loading station and the fortune; &quot; 'Out of the linear error correction operation step, the position storage step storage / = in the exposure domain The reference mark::, the flat method calculates the linear error correction for correcting the linear component error. The difference calculation step is stored in the storage step and is used to calculate the difference between the position of the exposure region reference mark in the light-area reference mark position. ; 193 200907596 The overlap control step is to use at least one of the expansion, rotation, and orthogonal methods obtained by the least squares method; In the difference, the exposure light is applied to the reticle s to calculate and control the overlapping target position of the projection image when the pattern image is projected on the exposure substrate and the exposure substrate. Eleven, circle: as the next page 194