TW200400540A - Projection optical system adjustment method, prediction method, evaluation method, adjustment method, exposure method, exposure device, recording medium, and device manufacturing method - Google Patents

Abstract

A wavefront aberration of a projection optical system is measured and information on the wave aberration is acquired (step 102). Moreover, a reticle pattern is transferred onto a wafer via the projection optical system (steps 104 to 108). Next, the wafer having the transferred pattern is developed, a line width of a resist image formed on the wafer is measured, and an image line width difference between the first line pattern extending in a predetermined direction and the second line pattern orthogonally intersecting the first line pattern is measured (steps 112 to 118). The projection optical system is adjusted so that the size of the ninth term (lower degree spherical aberration) is controlled according to the value of the twelfth term (higher degree astigmatism) of the Zernique multi-nominal in which the wave aberration is expanded and the aforementioned line width difference (steps 120 to 124).

Description

200400540 发明 Description of the invention: [Technical field to which the invention belongs] ▲ Nishimoto & Ming's method for adjusting the projection optical system, prediction method, flat method to 5 weeks, exposure method and exposure device, program and component system仏 Methods, step by step, in detail, it is a method of adjusting the projection optical system that projects the image on the first surface onto the second surface, and uses the projection light to the system to predict the characteristics of the 11 images. ㈣ This evaluation method is an evaluation method to evaluate the characteristics of the pattern image. Use this evaluation method to adjust = the adjustment method of the formation state of the case image. ㈣ This adjustment method or the aforementioned adjustment method of the projection optical system is very suitable for forming a pattern on an object. Implementation of the exposure method or the adjustment method of the projection optical system = an optical device, a process of causing a computer to execute the aforementioned prediction method < and a device manufacturing method using the aforementioned exposure method or exposure device. [Previous technology]-Generally speaking, "microfilm manufacturing" for manufacturing semiconductor devices, display devices, thin-film magnetic heads, microcomputers and other micro-components is made of hoods, which are formed on a hood or a reticle (hereinafter referred to as "standard" "Lines") are transferred to a sensing object (hereinafter referred to as a "wafer") such as a wafer or a glass plate through a projection optical system. The so-called stepper or so-called scanner (also called scanning Stepper) and other projection exposure devices. Known 'This exposure device measures the vertical line pattern of the optical system when measuring the line width difference between the vertical line pattern and horizontal line pattern transfer image (photoresist image, etc.) formed on the wafer by exposure. The reason why the contrast between the image and the horizontal line pattern image is 200400540 'is that asymmetric aberration such as coma aberration is considered as the main cause. Therefore, the measurement result is not easy when asymmetric aberration components such as coma aberration cannot be measured: positive line width difference. /… Recently, “In the assembly of the projection optical system,” the interferometer is used to measure the wavefront aberrations at various positions in the field of view (or exposure range) of the projection optical system. 'Zemike' polynomial (for example, Stripe checker polynomial) Expand the measured wavefront aberration (aberration function) series and adjust it so that the coefficients of each of the series (each checker term) of the obtained series (checker checker) Coefficients) are below the target values. This adjustment is performed because the aforementioned series of terms (each Chaney term) respectively indicate a specific wavefront aberration component, and each coefficient indicates the magnitude of each aberration component. Recently, the accuracy of aberration control of projection optical systems (projection lenses) has been improved by introducing wavefront aberration measurement into the above-mentioned process of the projection optical system, and it has been rapidly improved by using the expansion of the check polynomial of wavefront aberrations. . Also, according to the use of Zanek polynomials, the wavefront aberrations are divided into terms (each sine gram term) (the sine gram coefficient 彳 ^ ^ number), one ‘brother ’s number,’ and Zanek is inducted.

SensitlVlty) table linear combination to obtain the imaging performance of projection optics 'for example' aberration (or the index value), by the so-called "sensitometric sensitivity method" on the influence of aberration, can not be simple for the simple Way to judge. Here, the so-called Sensitivity Sensitivity Table refers to the exposure conditions, that is, the optical conditions (exposure wavelength, the maximum Na 1 setting, the opening shape of the opening of the illumination system, etc.), and the evaluation items (shade hood) (E.g., line width, evaluation amount, pattern information, etc.), the imaging performance of the projection optical system determined by the plurality of exposure conditions T specified by the combination of these optical conditions and evaluation items, respectively, 200,400,540, for example, various A calculation table consisting of the amount of change in each Zanek term of the aberration (or the index value). However, there is also an evaluation quantity that is not necessarily applicable to the so-called Zanek method. Regarding this line width change, as disclosed in page 713 of ν〇 · 4346, according to the rotationally symmetric component (0 0 component) of the aberration, and the second-order rotationally symmetric component ㈤ component), the line width becomes the maximum focus position change. , And the maximum value of its line width also changes. In addition, there are interactions between the two aberration components (the 2 component and the 2 component). Therefore, the so-called Chalk Sensitivity method is not applicable to the estimation of line width. The above-mentioned rotationally symmetric component (0Θ component) term that expands the wavefront aberration series using fringe check> 圼 gram polynomials includes a low-order term indicating defocus, that is, the fourth term (which is owed to ζ4) or low Of spherical aberration

These changes, the wave surface changes of the members are isotropic, so the effect of the pattern imaging state of the two (I line), Η line, and d line) is equal. Also, the second-order rotationally symmetric component (20-component) term has the fifth term (coefficient z5) representing low-order astigmatic aberration or the 12th term (coefficient zj) representing high-order astigmatic aberration, but for these 2Θ component terms The effect of the imaging state of the vertical line pattern: Although the sign of the image state is the opposite, the sign of the S wood is not equal to the size of the M. Therefore, it is not considered that the two components of the 1H00 component and the 2Θ component are not considered. (The coefficients (components) are all non-zero). ^ Line, k, line pattern ^ In this case, there is no simple and reliable $ 丨, see, for the line of longitudinal pattern and horizontal line pattern image. See. "The judgment method is difficult to adjust. 200400540 The first purpose of the present invention is to provide a method for adjusting the projection optical system under the above circumstances. In particular, it is capable of freely controlling mutually orthogonal line pattern images. The line width difference. '° The second purpose of the present invention is to provide a predictive method for i, nl +, and 丄 ^, which can easily and accurately predict the characteristics of the pattern image transmitted through the projection optical system. Brother 3 aims to use the ancestral site as an evaluation method for scholars and imaginary systems, which can easily and highly evaluate the characteristics of the pattern image through the projection optical system. Brother 4 of the present invention, your ancestor / i £ For the trimming method, it can easily and accurately adjust the characteristics of the pattern image transmitted through the projection optical system. The 5th purpose of the present invention is the ancestor / if μ. It is based on the previous method and exposure device. The system forms a pattern on an object. / The sixth object of the second invention is to provide a program that performs the prediction of the pattern transfer characteristics in a short time and with high accuracy through the projection optical system. The seventh object of the present invention is to provide The method of manufacturing components improves the productivity of components. [Abstract] At first glance, it will be considered that the rotationally symmetric component of the series expansion using the Charm's polynomial (e.g., ^ formula) In the (2Θ component) term, there is no correlation between benefits. However, the Mingming Temple repeats various experiments (including the memetic pupil plane (independent variable of the radial polynomial / upper's discovery and each component of the 2Θ component). The wavefront disorder with phase division of 1 θθ is sometimes different in the vertical and horizontal directions = 'in pupil plane'. For example, when 200400540 uses fringe lookup; ^ 圼 $ & ύ4τ, coefficient Z aberration level Item 12 of the number expansion ('2% injury is not zero') It is found that the size of the 9th a) of the spherical aberration component is changed by constituting the projection light and the movement and exchange of optical elements, thereby controlling the pupil Above and below the plane, there are 7 2 9 items (the coefficient distribution 'can adjust the line width difference of the aforementioned vertical and horizontal lines. The phase in the right direction of the work' is on the other side ... It is characterized by ::: projecting the pattern on ... Brother 1 process is to obtain the characteristics including the aforementioned projection optical system Information; Brother 1 Guangfu's 2nd process 'is the use of the second line extending in the predetermined direction configured on the aforementioned projection optics :::' ,: The image of the second line pattern orthogonal to Brother 1's surface, the first person of the mouth The line width of the first line width of the line pattern (which is the line width of the "i" measurement on the 2nd line of the "i" line, and the line width of the spring pattern image) and the line width of the second frame image); In addition, the line 3 process is based on the above-mentioned process characteristics and the aforementioned line width difference. In order to control the interaction between the above-mentioned optical characteristics and the aforementioned line width difference, 1 The ancestor has a small heart, adjust the aforementioned projection optical system. ,, said the $ 2 optical characteristics of the large == ^ 'can also use the projection optical system to form the 1st line pattern and the 2nd line figure, the first line The line width of the pattern image and ::: 丨 the 1st line width (see the line width of the brother line (which is the line width of the second line picture 10 200400540 case image)) can also form the first line pattern After measuring the image with the second line pattern, the line width difference is measured. If this is the case, for example, in order to control the size of the second light and poetry that is easy to adjust, adjust the projection optical system, so that the aforementioned γ can be controlled because of the i-th optical characteristic that is difficult to adjust the projection optical system. As a result, P is generated. Therefore, it is possible to freely and surely control the difference in line width between line patterns that are not easily orthogonal to each other. In this case, when the Zidian obtained by the current test 1 process is the wavefront aberration information of the projection optical system, in the aforementioned third process 2: using the Zernike polynomial, expand the aforementioned first and second waters. The number of complex grammatical terms m # paid to the wavefront aberration series variable is greater than 4th order: above P white (when the independence of the radial polynomial is not zero, according to the aforementioned second-order and second-order rotational symmetry In order to control the size and width of the component term, in order to control the size of the front ㉛2 # component term and the size of the aforementioned line-to-symmetry component term < the 成份 symmetric component term and the spin of the same order, in this case, the front: = mentioned projection optical system The 2nd item of the 0 component item, the front and ^ rotationally symmetric component items are the 9th item of the 4th order C0S2 item, or the aforementioned ^ f rotationally symmetric component item is the 13th item of the 4th order 0Θ component item, the aforementioned spin two The% symmetric component term is the fourth-order sin2e into the ninth term. 疋 The symmetrical component term is the fourth-order θθ component term of the invention according to the projection optics of the present invention. The information obtained is the projection optics, and the first adjustment method is When the first process of the first process is to directly measure the wavefront aberration information of the system j, With reference to the aforementioned wavefront aberration information, the wavefront of the H-shadow optical system can be used to obtain or describe the first line pattern (which is different in configuration) of the first process based on the 200400540 array in each group. According to the measurement result, the best level of the two-line pattern on the / brother surface is ^^ / 0. According to the measurement result, the difference between the &, +, and the focal position can be estimated, and the root is A2. , +, + The information of the second-order rotationally symmetric component term is used as the information of the wavefront aberration. 乂 刀 贝 （贝 λ λ The information obtained by the first projection optical system of the present invention ... with the method of 1 The 1st system, 10th, and 1st are the cost of wavefront aberration of the shirt optical system. ^ The 3rd process of the brother is the second-order rotation ~ hole 訏, then the size of the aforementioned "process ..." is not zero. In addition, the above-mentioned line of the second-order rotationally symmetric composition of the aforementioned line is designed according to the size of the wind-cutting member and the width of the front machine *. The width approaches the design value, in order to make the line The rotationally symmetric component term is the same as H \ Yu, and first 'to make the aforementioned 2nd order. Η ρ The rotationally symmetric component term of the white number In the i-th projection optical system with the best size of the present invention, through the projection optical system, the first !, the second =: Brother 2 process projection image) is formed on the second surface of the line image (interim image, can also be Calculate the first and second image detectors to measure these empty pp ^ The first line pattern and the second line pattern image of the younger brother are continuation * limited to this, the aforementioned second process system includes: see the line of rare images' but Unlike the forming process, Chuanpi ## The body arranged on the second surface mentioned above, the image of the second line pattern is formed on the IW-straight object; and the line width measurement process, which is the first one described in the first measurement The line graph is on the object-, ^ the line width of the image), and the second line width (forms the above-mentioned second line pattern on the body is formed in the object-level system or SEM, etc., Lai Dan, | The line width can be obtained by measuring the latent image, the first blocked image, or the etched image of the first line and second line pattern formed on the object with the tone of the exposure device 2 绐 m. 12 200400540 Manufacturing invention ... The adjustment method of the projection optical system, in the third aspect, at least one of the position control of at least the direction of freedom constituting the aforementioned projection optical system and the control of gas acoustic force in a part of the optical path is provided. Control method of the aforementioned second optical feature "factory of the present invention. The optical system adjustment method of the present invention is described in the foregoing; M: the case is the vertical line pattern 'the aforementioned 帛 2 line pattern is the horizontal line pattern, the aforementioned i = characteristics and the second The optical characteristics are obtained through the process of obtaining the Chanck sensitivity in the cross term of the combination of the Zernike term combinations of the line width changes of the aforementioned vertical line pattern image and the aforementioned stern line pattern image, and the cross term The buried sensitivity is determined by the manufacturing process of the combination of the checked and buried items that are different in the vertical and horizontal lines. The first projection optical system adjustment method of the present invention, the aforementioned == received information is the aforementioned projection optical system. Information on wavefront aberrations:: The first and second optical characteristics are described by using Chase polynomials, which are the same order in the plurality of Chainec terms that expand the wavefront aberration series obtained by the aforementioned brother / process. And kind The term of different components. From the second point of view, according to the second exposure method of the present invention, a circuit pattern on the first surface is transferred to an object arranged on the second surface through a projection optical system. Including: an adjustment process, which uses the method for adjusting the projection optical system of the present invention to adjust the aforementioned projection optical system; and a transfer process, which uses the aforementioned adjusted projection optical system to transfer the aforementioned circuit pattern On the aforementioned object. If so, because the projection optical system is adjusted using the method of adjusting the projection optical system of the present invention, in order to make the line width difference between the vertical line pattern and the horizontal line 13 200400540 pattern image become a design Value, adjust the projection optical system. For example, adjust the projection optical system so that the line width difference between the vertical line pattern and the horizontal line pattern image of the same line width is the smallest (for example, zero). In addition, the projection optical system using this adjustment, System, so that the circuit pattern is transferred to the object, so that the transfer of high-precision patterns that reduce the line width difference between the vertical line pattern and the horizontal line pattern can be realized. From the third point of view Then, the i-th exposure dress of the present invention is to transfer the pattern formed on the hood on the object through the optical system for exposure. It is characterized by having an i-th projection optical system using the present invention. The adjustment method to adjust the method. I have to learn the system and $ μ > + η κ t shirt first to learn the system as the aforementioned exposure optical system. ', If this is the case, because the use of the first projection of the invention Optical system adjustment method to adjust the projection optical system as the optical system for exposure: 'So using this projection optical system, the pattern formed on the hood is transferred to the body' #This can reduce the vertical line pattern and The transfer of the precision pattern of the horizontal line graph. From the point of view of the fourth aspect, the present invention is a method of adjusting 2 and is a method of projecting a pattern image on the first surface onto the 21st: It is characterized by including: W 上 '具The first system is to obtain the information including the aforementioned investment characteristics; and ... the younger brother of the system 丨 the second optical process is based on the values of the i system characteristics and the line pattern (in the direction of the aforementioned light setting) (Extended) line width and 2nd line pattern-set Both the green and the eve of the night ^ ”draw the difference between the 1 line pattern of the fans, in order to control the interaction with the first optical characteristics 14 200400540 The first line pattern formed on the second surface is like the line width 2 : The second optical characteristic in which the line width difference of the pattern image line width is affected: Large J adjusts the Lishu projection optical system. If this is the case, then according to the difference between the optical characteristics of the first and second line patterns and the line width of the second line pattern, the $ 1 control line graph is% 1 7 ^ ^ ^ α The interaction between the two lines is different from that of the two lines. It is formed by projecting the ancestor system and formed on the second surface of the official disc II, +, and M. The line pattern image of the line 1 and the line 2 The difference between the line width of the 2-line pattern is too large, and the projection optical system is adjusted due to the & ^ difference).

^ ^ k, the difference between the line width of the 1st line pattern image on the 2nd line and the 2nd line pattern line image difference (line width difference), because the 1st line pattern line width and the 2nd line on the 1st line When the difference between the pattern line widths occurs, for example, 'if it is caused by a day-to-day error in the pattern on the hood, the line width difference between the orthogonal line patterns can be freely controlled. In this case, the first line pattern is a vertical line pattern, the second line pattern is a horizontal line pattern, and the optical characteristics and g

Based on the cross term of the combination of the Zanek terms of the line width image and the line pattern image, the dragon ψ is left, θ, and the process of the Zanek sensitivity is calculated and calculated. The process of combining the Zanek sensitivity symbols in the other terms with the Zanek terms of which the vertical and horizontal lines are different is determined. From a fifth point of view, the second exposure method of the present invention uses a projection optical system to transfer a circuit pattern on the first surface onto an object arranged on the second surface, and is characterized by including: : The adjustment process is to adjust the aforementioned projection optical system using the adjustment method of the second projection optical system of the present invention; and 15 200400540 optical system, the front transfer process uses the aforementioned adjusted projection circuit pattern transfer On the aforementioned object. 2. If this is the case, because of the production system, the method of adjusting the projection optical system is to complete the projection optical system and system for 5 weeks, for example, due to the daylight error of the pattern on the hood: When the difference between the line width of the first line pattern image and the line width of the second line pattern (line width difference) on the second surface is adjusted, the projection optical system is adjusted so that the orthogonal line patterns can be freely controlled. Linewidth difference 'due to the projection optical system using this adjustment

Transferring the circuit pattern to the object now reduces the high-precision pattern transfer between the vertical line pattern and the horizontal line pattern. From the point of view of the sixth aspect, the second exposure device of the present invention transfers the pattern formed on the hood to the object through the optical system and system for exposure. It is characterized by having: The second method of adjusting the projection optical system of the invention is to adjust the projection optical system of | as the aforementioned exposure system. If this is the case, a projection optical system adjusted using the second projection optical system adjustment method of the present invention is provided as an exposure optical system. Therefore, the projection optical system and system are used to transform the pattern formed on the hood. It can be printed on the object, which can realize the transfer of the pattern with high precision ^ which reduces the difference between the width of the vertical line pattern and the horizontal line pattern. From the seventh point of view, the adjustment method of the third projection optical system of the present invention is to project the pattern image on the first surface onto the second surface, which is characterized by including: obtaining the aforementioned projection optical system The process of wavefront aberration information; 16 200400540 The process of obtaining the projection image information about the aforementioned pattern; and the process of aging, which uses the 杳 series to expand a plurality of 项 μ terms = Zhongxi ': in the aforementioned wavefront aberration The crossover of any combination of grammar terms affected by the projection image characteristics = ΐΐΓ: The change in sensitivity to the aforementioned projection image characteristics to adjust the aforementioned projections. If this is the case, a projection optical system can be obtained, and the relevant map can be obtained. After the death of the horse, "The aberration case of the aberration", like the information. And, based on this information

", When the projection optical system is used", the Chase polynomial is used; · The complex number, J, which will expand the wavefront aberration series, affects the aforementioned projection image characteristics: The mutual six ▽ linguist is also an example of a combination of Renschanek terms, taking into account the change in the characteristics of the projection image by the Channeck sensitivity, and adjusting the aforementioned projection optical system. That is to say, the adjustment characteristics of the sub-Γ's Yihuakou ^, ^ Λ in the 3rd and 4th optical system of the M °°, whose interactions are not considered are at the intersection of the combination of arbitrary search terms that make the image of the pattern ... Gram = changes in frontal image characteristics, adjust the aforementioned projection light; system: system = aberration components that are difficult to ride, for example, high-order aberrations can also be adjusted ^ the projection optical system can be adjusted to make the pattern image formation better. In case k, when the aforementioned pattern includes a line pattern, the characteristic of the aforementioned projected image includes the line width of the line pattern. … If viewed from the eighth point of view, the third exposure method of the present invention is to transfer a circuit pattern on the first surface onto an object arranged on the second surface through an optical system, which is characterized in that: Including: 17 200400540 The adjustment process 'adjusts using the third projection optical system of the present invention' method to adjust the aforementioned projection optical system; and, the transfer process uses the adjusted projection optical system to convert the aforementioned circuit pattern Transfer on the aforementioned object. < If so, because the third projection optical system adjustment method of the present invention is used to adjust the aforementioned projection optical system, the projection optical system is adjusted: 'to make the formation state of the pattern image better, because this adjustment is used The projection optical system transfers the circuit pattern to the object, so it can transfer the pattern with high precision. From the ninth point of view, the third exposure device of the present invention is a pattern that is formed on a hood and is transferred to an object through an exposure optical system. It is characterized in that: the present invention is used: The projection optical system adjusted by the third projection optical system is used as the aforementioned exposure optical system. ^ In accordance with the present invention, since a projection optical system adjusted using the third projection optical adjustment method of the present invention is provided as the aforementioned exposure optical system, the optical system is used to transfer the pattern formed on the hood # On the object, this can achieve high-precision pattern transfer. From the tenth point of view, the present invention is a fourth exposure device. The beam is used to illuminate the pattern arranged on the first surface, and the aforementioned pattern is transferred to the second surface through the projection optical system. On the object, it is characterized by having: ... an optical characteristic measuring device for measuring optical characteristics (including the first optical characteristic of the aforementioned projection optical system); 18 200400540 The line width measuring device is for measuring the i The difference between the line width of the line pattern (which extends in the predetermined direction configured on the aforementioned =) and the line width of the second line pattern (which is orthogonal to the first line pattern described above); the image formation state adjustment device ' It is used to adjust the formation state of the aforementioned projection image, and the first position is based on the value of the %% characteristic measured before the optical characteristic measuring device and the first line width (using the aforementioned line width measuring device The measured line width of the aforementioned first line pattern image is 2 and the width of the line pattern image is measured. The line width (because of the aforementioned second suppression, the use of the aforementioned image formation state adjustment device = 2 first learning characteristics (because of And the above-mentioned optical characteristics, so that the line width difference is affected If the effect is in accordance with the present invention, it can be measured by the line width measuring device, including the projection optical system, and the measurement system, respectively, on the second side ( The image plane and pattern (which are the i-th plane (the object extends in the same direction) and the second line graph formed by the projection light on the first plane (image plane) are here. Here, the line width measurement setup is used. (Line width, pattern orthogonal) The same-inside measurement on the objects arranged on 2 sides can also measure the transfer image (latent space) of the line width horizontal line pattern formed on the first image, photoresistance image, _image) The image is formed on the second surface. Let ’s say: the vertical line pattern and the horizontal line pattern are controlled again. When the line width of the optical space image is used. When the scientific characteristics exist, it is based on the first installation: the amount: ... Is the difference (line width difference) between the sum of the first line width (19 1 line width (the second line pattern == the line width of the pattern image) and the line of Takarazuka) measured by the line width measuring device, Using the pico ^ UU4UU540 as a state adjustment device, the line width difference is affected by the optical characteristics of the brother 1 to make Le 2 learn the size of the characteristic first. Therefore, for example, / For example, it is easy to adjust the optical characteristics of the 2nd optical characteristics that are difficult to adjust for the first 2nd time! ::: Image formation state adjustment device to control the size of the 2 prior characteristics for control Control 帛! The aforementioned line width difference caused by the existence of optical characteristics. Projecting two or two energy beams to illuminate the pattern arranged on the first surface, and through the characteristics, it can achieve the right 4 access to the object arranged on the second surface The optical characteristics measurement device of Liangzhou Image 2 which measures the line width difference between the intersecting line pattern transfer images is to measure the aforementioned wavefront aberration measurement device. Next, the aforementioned first optical characteristic is developed by using the Buick polynomial system to measure the wavefront aberration measurement of τ Kexi ’s Zanikex order 4th order and the wavefront aberration series of Riley. The second optical characteristic is the aforementioned second order. :: = rotationally symmetric component term, front rotationally symmetric component term. In this case, the term of the symmetry component is the same as that of the second order component. Any of the terms 12 and 13 of the second order component is a 4th order 2Θ, which is the 9th item of the 4th order. The aforesaid rotationally symmetric component item of the fourth exposure device aerial image measuring device of the present invention measures a projection image which can also be included in a cutting width-measuring device, and the aforementioned respective pattern settings on the two sides of the line width measuring device It can also include a photographing device. It takes 20 200400540 to take an image of an object formed on the second surface. The fourth exposure nightmare of the present invention, a, > ^ ^ Second, the image formation state adjustment device is located at least i, at least i of one optical element, and = degree =, which constitute the aforementioned projection optical system. s weekly adjustment, adjustment of gas force in part of the optical path, the adjustment of the wavelength offset of the energy beam, and the aforementioned projection light of at least one of the pattern forming member and the aforementioned object regarding the aforementioned pattern: In the adjustment of the optical axis direction of the system, at least one adjustment is performed. From the perspective of the Uth point, this is a prediction method that predicts the characteristics of the pattern image by investing in an optical system. It is characterized by including the prediction = Cheng 'system, and according to the plural terms (including the formula using ㈣ , The aberration components obtained by expanding the wavefront aberration series of the above-mentioned four systems are calculated. Under the given exposure conditions, the variation curve is calculated (representing the predetermined projection through the projection optical system). The image of the pattern, the size of the aforementioned image, and the variation of the defocus amount from the optimal focus position) are used to predict the aforementioned variation curve based on the calculated movement. When transferring a pattern through a projection optical system, you can understand the pattern: Image: The size changes according to the amount of defocus from the optimal focus position of the transfer position, which indicates the change curve of the change (that is, the so-called CD _Focus Curve) changes due to wavefront aberrations of the projection optical system. In addition, the wavefront aberration of the projection optical system uses a predetermined formula (for example, Zernike polynomial) to further expand the series. 11 This can be decomposed in a plurality of search terms (aberration component terms). Based on the results of intensive research by the inventors, it can be understood that the values of the linear combination of the plural terms including the coefficients (ie, aberration components) of the above-mentioned check / gram terms, 21 200400540, and the image of the pattern projected through the projection optical system Kization of the aforementioned change curve (g 卩 'regard the defocus amount and the size of the pattern image as the coordinate system of each coordinate axis, the direction of the political focus and the parallel movement of the change curve of the image size direction)' There is a close relationship.

Say 'If you rely on the prediction method of the present invention, the above relationship is used.' It is not necessary to use the imaging branch caused by complex calculations that require a lot of calculation time. ^ λ, the value of 、, Ό, '' can be used to predict the focal curve of the pattern in a short time under the projection optical system (under a predetermined exposure condition of aberration), and according to the prediction result, Use a short time to predict the characteristics of the projected image (or transfer image) of the pattern. In this case, 'can further include the variation curve obtained; a process like a high-order function, which is described before, before the prediction process, In the case of front exposure: assuming that there is no aberration in the projection optical system, the precision curve j is obtained by simulation to indicate the change curve of the size of the image to the defocus amount.

In the case of money, the aforementioned forecasting process is based on the aforementioned aberrations under the predetermined exposure conditions, and the aforementioned aberration components are invited to the aforementioned: The coefficients are linearly combined to calculate the amount of movement in the direction of the two joules described above. According to the aforementioned aberration components (which are in small change and less, as soon as the square of each aberration is injured), Sensitivity is used as the coefficients) and the amount of movement in the direction of image size change is described in the square curve. As mentioned above, the movement of the 'variation curve' can be decomposed into the relevant defocus amount 22

20040U54U The axis direction (the direction of the defocus amount axis) curve ί 10 ^ + movement, and the relevant representation of the figure Zai Fool small axis direction (like the direction of the large and small axes) 敕 釭 Mouth wood image large m ^ Λ moves. The movement of the k-motion curve in the direction of the defocus amount axis is the time when the wavefront aberration of the system is given in advance.

Among the aberration components, there is sensitivity, and the amount of movement can be predicted by the linear combination of the D, 丨 丨 I # 妙 4, 彳 tsuka difference components. In addition, the movement of the curve of the axial direction of the image is reduced to the square of each aberration component, and there is a snoring sound. — Sand movement is called by people to the degree of error by the square of each aberration component. Sex ... d 'can predict this amount of movement. w In the case of a spoon, the aforementioned forecasting process is based on the foregoing terms (including the line of the square of each aberration component (known as Niu II: the direction of the change in the size of the aforementioned image under the given exposure conditions) = this The difference between the different aberration components and the sensitivity of each item is taken as the shift of each change: the second combination is to close the aforementioned image size of the aforementioned change curve, and the change in the direction of the magnitude axis of the image. Moving, there is sensitivity not only in the aberrations of various aberrations, but also in the intersection of aberration components that are different from each other. There is also a sensitivity 'If you further consider the linear combination of these intersections, you can enter- The amount of movement in the direction of the image axis is accurately predicted step by step. In the prediction method described in the item, the aforementioned high-order function is a function composed of even-order terms only. In the prediction method of the present invention, the aforementioned prediction process is based on The linear combination of the plural terms of the aberration components is calculated to calculate the aforementioned variation curve, and the deformation caused by the aforementioned wavefront aberration can be predicted according to the aforementioned movement amount and the aforementioned deformation. Measure the aforementioned variation curve. In this case, only the amount of movement of the variation curve can be calculated ', and according to the linear combination of each aberration component term 23 200400540, the deformation of the wavefront aberration variation material of the projection optical system can also be calculated, which can be more Predict the variation curve with high accuracy. In each case, the variation curve obtained in the aforementioned prediction process seems to be broken *, and the ratio is suitable to include a process that approximates the X-motion curve to a order function, which is at 2. ], It is assumed that the aforesaid projection optical system has no aberration 2 and is calculated by simulation to indicate that the aforementioned image is underdeveloped and wrong. 31 The variation of the size of the image with respect to the variation in the defocus amount is different in this case. Before the manufacturing process, further package A has been installed, which is through the aforementioned projection optical system of the actual aberration state = under the given exposure conditions, to calculate the fortunate image of the aforementioned image being projected = the variation of the variation curve Γ 屮… The M function (which is calculated according to the aforementioned moving system r) and the difference function (which represents the difference of the curve function obtained by calculating I ^ as described above). In this case The aforementioned calculation process is performed by simulation. Aberration: Γ Mingzhi: The measurement method is to obtain a higher-order function (in the aforementioned movement amount) in the case of the aforementioned wavefront ε: variation of the other curve in the prediction process. When the movement curve is approximated) and the difference between the movement curve function obtained by the calculation process described above), (2 == each of the aberration components (under the predetermined exposure bar, the aberration components will be described below) The square of the function of the difference is regarded as the coefficients. The linear combination of the squares is calculated to calculate the coefficients with the aforementioned ^ the first order term. According to the aforementioned aberration components (under the conditions of 疋 a above, the aforementioned aberrations are calculated). The sensitivity of the component to the aforementioned difference function is the linear combination of the 24th of the 200400540 order terms as the coefficients) to calculate the coefficient of the odd order term of the aforementioned difference function%. In this case, it represents the deformation of the variation function. The coefficients of the even-order terms of the difference function have sensitivity in the square of each aberration component when the wavefront aberration of the projection optical system is expanded, and can be predicted by the linear combination of the squares of each aberration component. Number. The coefficients of the odd-order terms of the difference function have sensitivity in each aberration component, and the coefficient can be predicted by the linear combination of the aberration components. Therefore, the deformation of the variation curve can be predicted in a short time and with high accuracy by using a linear combination of aberration component terms including wavefront aberrations of the projection optical system. _ In the prediction method of the present invention, the aforementioned predetermined formula is a Zernike polynomial, and the aforementioned aberration components are coefficients of each Zernike term. From the perspective of the twelfth aspect, the present invention is the first evaluation method, which evaluates the characteristics of a pattern image through a projection optical system, and is characterized in that it includes

The prediction process is based on 40 pairs of the aforementioned projection optics and at least one measurement point within the effective field of view. Using the method of the present invention, under a given exposure condition, the variation curve is predicted through the aforementioned projection optics (system Represents the predetermined pattern image projected on the at least one measurement point, and the change of the image size to the defocus amount from the optimal focus position); and the characteristics of the evaluation process image. It is based on the aforementioned prediction results to evaluate the predetermined map according to the first evaluation method. At least one measurement point in the effective field of view. The line (for the prediction using the present invention can be aimed at the high performance of the projection optical system. Precisely predicting the above-mentioned variation curve method 'under a given exposure condition, the image of a predetermined pattern projected by a projection optical system is transmitted through 25 200400540. Therefore, based on this variation curve, it is possible to accurately evaluate the effective vision of the projection optical system ... . In this case, the predetermined pattern is arranged corresponding to the plurality of measurement points in the effective field of view of the projection optical system, and the characteristic spoon and the effective field of view of the projection optical system are included. The uniformity of the aforementioned image / and the first evaluation method of the present invention, the predetermined pattern includes = :: a plane orthogonal to the optical axis direction of the projection optical system-two orthogonal to each other-the aforementioned prediction process performance The above-mentioned variation curve. Q. Prediction In this case, the aforementioned evaluation method can evaluate the difference in line width between line pattern images with respect to the characteristics of the aforementioned image. In this case, in terms of true characteristics, for example, it can be evaluated that the main cause is astigmatism: the difference between the line widths of the two line patterns of the father of Jin Cheng. As for the first evaluation system of the present invention and the above-mentioned system :: === Case 2 ::::: 'The aforementioned predecessor system can be shown in a picture. In this case, the aforementioned evaluation system of the blind monk #U ”贝"The king can evaluate the aforementioned image characteristics of the line pattern images. In this case, it is possible to evaluate the line width anomaly value mainly due to the aberration 4 as the aforementioned image characteristics. -Adjustment = 2? From the point of view, the present invention is the first adjustment method, which is the formation state of the pattern image of the optical system. It is characterized by 26 200400540, and the manufacturing process uses the first evaluation method of the present invention to evaluate the corresponding description of the cross shadow optical The characteristics of the predetermined pattern image configured by at least one measurement point in the effective field of view of the system; and the manufacturing process adjusts the formation state of the predetermined pattern image through the projection optical system according to the evaluation result. If the first adjustment method is based on the right, then using the first evaluation method of the present invention, the existing case image feature of at least one measurement point in the effective field of view of the Khanbei projection optical system is based on the evaluation result, through the foregoing The projection optical system adjusts the formation state of a predetermined pattern image. Therefore, based on the evaluation results, it is possible to adjust the characteristics of a predetermined pattern image to a desired state. In the case of a variety, the aforementioned adjustment process uses an amount of change of the aforementioned aberration components per unit adjustment amount of the adjustment parameter, which adjusts the formation state of the aforementioned predetermined pattern image about the aforementioned measurement points. 9 The aforementioned aberration components Sensitivity to the change in the size of the aforementioned predetermined pattern image; and the deviation of the target values of the coefficients of the respective order terms of the relevant curve of X ', which is not to describe the change of the size of the predetermined pattern image to the aforementioned defocus amount; The adjustment amount of the 5-week integer parameter is described, and the formation state of the predetermined pattern image is adjusted according to the calculated adjustment amount. The change curve of the measurement point is affected by the aberration of the projection optical system. Therefore, if the projection optical system is adjusted so that the aberration component changes, the variation curve of the measurement point can be brought closer to the desired curve (target). Here, the present invention uses the adjustment amount of 27 200400540; the change of the aforementioned aberration components per unit adjustment amount of the parameter Under the aforementioned predetermined exposure conditions, the sensitivity of the aforementioned respective aberration components to changes in the aforementioned predetermined pattern image size; related The deviation of the target value of the coefficient of each order of the variation curve represents the variation of the aforementioned predetermined pattern image size to the aforementioned defocus amount; / The adjustment parameters required for the deviation of the variation curve of the measurement point from the desired curve (系 调整 图 图 傻傻 夕 # # Adjustment of the formation state of the danmen 1 mouth tea image) (adjustment parameters) ® 'According to the calculated adjustment amount, pinch Pu door 4 Formation of evil. Hunting this, the pattern image and the mouth-shaped image can be adjusted to make the shape / evil of the image, so that the size of the pattern image, the ",": E curve close to the desired curve. This clear shape 'will What kind of curve is formed by the desired curve (brief 7 b M mesh "), the adjustment items of the pattern image obtained due to the weight are eight seconds, 4 fine, for example, the aforementioned evaluation process, and the knife do not correspond to each of the foregoing Projection optics Li Gedan, Japanese I ^ / Wu Yu 糸 system of the fixed number of points arranged in the effective field of view of the predetermined map 荦 silly related to the characteristics of the front m 前述 image, the aforementioned adjustment process is to make the Guan The target values of the phases π, ', in the same order are at the aforementioned measurement points. In this case, the in-plane uniformity of the circular image in the effective field of view of the first learning system can be improved. In addition, when the two, +, and Uighur patterns contain a plurality of patterns, the moon dagger is between the patterns, so that the items of the first order include the same order of the items of the 1st verbal motion curve ^ value. . In this case ’, for example, the seat is like a toilet;); to green 罔 also commits to make the vertical line pattern lesson as flat as possible with the line width or flat line pattern. Adjusting the line widths of the fixed images, etc. The first adjustment method of the present invention uses an adjustment amount. Take the Xiaoping method to find the aforementioned 28 200400540. If you look at the point beam, the present invention is the fourth exposure method. Through the projection optical system, the circuit pattern on the 4th surface of the flute on the flute is transferred to the The object arranged on the first surface is characterized in that it includes: an adjustment process, which uses the " rounding method of the present invention to project the optical system, adjust the formation state of the aforementioned pattern image, and projection == ,: : = In the state where the whole image is formed, — 予 =, first transfer the circuit pattern described above to the aforementioned object. Therefore, according to the exposure method of the 4th, the first adjustment of the shadow optical system of the present invention is used to adjust the formation state of the circuit pattern image. In the formation state of the f adjustment image, the circuit pattern can be transferred to the object because the circuit pattern is transferred to the object. This circuit pattern is formed on an object with high accuracy. From the perspective of the fifteenth aspect, the present invention is the first method to evaluate the characteristics of the pattern image through the projection optical system, which is characterized by: 'the process of obtaining the wavefront aberration information of the aforementioned projection optical system; The process of projecting the projection image information of the pattern; and the evaluation process is to consider the sensation sensitivity to the aforementioned projection image (which uses the Zernike polynomial 'to expand the wavefront aberration series of the plurality of Zernike terms, which The change in the characteristics of the projection image of the cross-term of any combination of cross-term terms that affects the characteristics of the projection image to evaluate the aforementioned pattern image characteristics. According to the second evaluation method, the wavefront aberration information of the projection optical system can be obtained, and further information about the pattern projection image can be obtained. In addition, according to these information, 'in the use of Chasing grams to expand the aforementioned wavefront aberration series to a plurality of 29 Chawkin terms', at the intersection of the combination of the desired Chaniek terms, L and the influence of image characteristics In X Bei, the changes in the characteristics of the ancient images of 杳, g are examined, and the _ Xinyi / Rick sensitivity of the aforementioned projection evaluation method, two: the characteristics of the Γ image. That is to say, in the case of arbitrary investigations on the influence of the second characteristic, the sensitivity of the aforementioned projection image to the aforementioned projection is # # Μ π and the member, because of the consideration of Zanek, it is possible to use a different pattern image. The characteristics of the moon dagger are more accurate than the characteristics of the pattern image.

In this case, when the aforementioned R denier A j4 pattern includes a line pattern, the aforementioned image characteristics include the line width of the line pattern. From the 16th point of view, 4 μk, # 坰 敕 ', S, the second adjustment method of the present invention, which passes through the projection optical system, includes: 3 cases like M Si, which is characterized by- The manufacturing process is to use the second evaluation method of the present invention to evaluate the characteristics of a predetermined pattern image assigned to at least one measurement point in the effective field of view of the Qianying optical system;

System to adjust the shape of the former Qianding pattern image. ... If the second dimming method is used, the second evaluation method of the present invention can be used to accurately evaluate the predetermined pattern image characteristics of at least one measurement point corresponding to the effective field of view of the projection optical system. As a result of the evaluation, the formation state of the predetermined pattern image was adjusted through the projection optical system. Therefore, based on the evaluation results, the formation state of the pattern image can be optimally adjusted. From a 17th point of view, the present invention is the fifth exposure method. The pattern on the first surface is transferred to an object arranged on the second surface 30 200400540 through a projection optical system. Its characteristics It consists of: The adjustment process is based on the application of the patent _ the first through the projection optics ", the formation of the aforementioned pattern image ^ method and the 'transfer process, is based on the shape of the aforementioned adjustment image' through projection: learning system, the aforementioned The pattern is transferred on the aforementioned object. Let alone, from the perspective of Brother 18, the present invention is a computer through a projection optical system " is characterized by the prediction of the characteristics of the case image, its wavefront = The linear combination of a plurality of terms (including the aberration components obtained by expanding the mouth of the aforementioned projection optical system by a series using a predetermined formula), and under the existing exposure conditions, the predetermined pattern image of the aforementioned projection is used to calculate Regarding the focal imitation department M: the cause of the aforementioned wavefront aberration of the variation curve of the size of the described image to the variation from the aforementioned best focus shift amount of political focus, the aforementioned computer is moved according to the aforementioned The amount of movement is used to predict the prediction procedure of the aforementioned variation curve. ~ As long as the program is installed in a computer, the computer executes the above-mentioned procedures. = This' The prediction method of the present invention is executed by a computer. In this way, instead of using the complicated calculations that require a large amount of calculation time to mimic 'through a very simple operation (to find the line I · generation combination values of the complex members that each contain aberration components)' The projection optics of the aberration state can be used to predict the relevant CD-curve curve in a short time under the exposure conditions. Based on the results, the pattern transfer characteristics can be predicted in a short time. In this case, before the aforementioned prediction procedure, it is assumed that the aforementioned projection 31 200400540: Learning: When there is no aberration, 'the computer is used to further execute the variation curve (:: the change in the image size to the aforementioned defocus amount). In the case of the high-order function t, the aforementioned prediction program is a program that uses the aforementioned computer. It is based on the aforementioned aberration components (which are the front-speed images under the predetermined exposure conditions). As a component of each coefficient) of the line of the defocus amount of the sensitivity dispersion

Procedure for predicting the shift of the defocus amount direction of the aforementioned variation curve; The sensitivity of change is treated as a linear combination of coefficients and squares to predict the amount of movement in the direction of change of the aforementioned image size of the aforementioned change curve. In the program of the present invention, the aforementioned prediction program causes the aforementioned computer to execute the following procedure, which includes the linear combination of the squares of the aberration components described above. difference

It is a procedure for predicting the amount of movement of the image size change direction of the aforementioned change curve by linear combination of the changes in the size of the image just described as each coefficient). In the program of the present invention, the aforementioned higher-order function is a function composed of only even-order terms. In the program according to the present invention, the aforementioned prediction program is based on the linear combination of the plural terms of the aberration components respectively included to calculate the deformation state of the variation curve due to the wavefront aberration described above. To make the computer execute the program for predicting the aforementioned variable curve. In this I " monthly condition, before the aforementioned prediction procedure, the computer further executes a calculation procedure, which is obtained under the aforementioned predetermined exposure conditions, and false = the projection optical system is obtained without aberration, so that the aforementioned The variation curve of the small variation of the aforementioned defocus amount approaches a higher-order function. In the case of the attack, before the aforementioned prediction step, the brain is further used to execute the calculation program, which is the aforementioned projection optical system under the above-mentioned light condition, = the state of the transversal aberration, to calculate the projection related Pattern image 'The variation of the aforementioned image size with respect to the aforementioned defocus amount. The prediction program uses the computer to execute the difference function of obtaining the higher-order function data and the moving amount) and the variation function (which is obtained by calculating from the above calculation steps) as the difference of the curve caused by the wavefront aberration. Procedures for changes. In this case, the aforementioned prediction program is a program executed by the aforementioned computer, and based on the respective aberration components (under the predetermined exposure conditions, the sensitivity of the square of each aberration component to the even-order term of the difference function is described). A procedure for predicting the coefficient of the even-order term of the aforementioned difference function when the linear combination of the squares of the two respective coefficients); and according to the aforementioned aberration components (under the aforementioned predetermined exposure conditions, A procedure for predicting the coefficients of the odd-order terms of the aforementioned difference function as a linear combination of the sensitivity of the odd-order terms of the aforementioned difference function as each coefficient). '$ The formula of the present invention The aforementioned formula is a Zernike polynomial, and the aforementioned 33 200400540 each aberration component is a coefficient of each Zernike term. The program of the present invention can be used as an object for sale in a state of being recorded on an information recording medium. Therefore, if viewed from point 191, the program of the present invention is an information recording medium read by a recorded computer. 4. Furthermore, from the perspective of the 20th aspect, the present invention relates to a method for manufacturing an exposure device. The invention relates to a method for manufacturing an exposure device. A projection optical system is used to transfer a pattern in the shape of a hood on an object. Use of the invention

Any one of the first to third methods of adjusting the projection optical system & adjusts the manufacturing process of the aforementioned projection optical system. In the photolithography process, exposure is performed using any of the fourth to fourth exposure apparatuses 2 of the present invention, and a pattern can be formed on an object with high accuracy, thereby making it possible to produce a higher integration rate with a high yield. Micro components can improve their productivity. Similarly, in the lithography process, exposure is performed by using any one of the fifth to fifth exposure methods of the present invention—the method can be used to form a pattern on an object with high accuracy, thereby making it possible to produce a higher product with a high yield. Body microelements,

Can improve its productivity. Therefore, from another point of view, the present invention is a device manufacturing method using any one of the first to fourth exposure devices of the present invention, or a device manufacturing using any one of the first to fifth exposure methods of the present invention. method. [Embodiment] Hereinafter, one embodiment of the present invention will be described with reference to Figs. 1 to 12. Fig. 1 shows an outline of an exposure apparatus according to an embodiment. 34 200400540 Kouhai exposure equipment 1GG series, exposure light source (hereinafter, referred to as "light source") + 'step scanning using a pulsed laser light source Method of reducing the projection exposure device (so-called scanner). The exposure device 100 is provided with: an illumination system (which is composed of a light source 16 and an illumination optical system 12), a reticle stage RST (which is used as a hood stage 'retains a reticle 纟 R that serves as a hood, By the energy from this lighting system: the exposure of the beam is illuminated by the illumination light EL), the projection optical system pL (which is the exposure illumination light el emitted from the reticle R is projected on the object as the object Circle (image plane)), wafer stage WST (holding the crystal mesh w), and these control systems. For the aforementioned light source 16, a KrF excimer laser (with an output wavelength of 248 nm) is used here. Also, copybook, shield! & Second, t, especially the original 16 and Lv, can also use the pulsed UV light source in the vacuum ultraviolet field such as output F2 laser (output wavelength 157nm) or ArF 帛 molecular laser (output wavelength 193nm). The above-mentioned light source 16 is actually provided with a room u for housing the exposure device (consisting of the constituent elements of the illumination optical system 12 and the reticle stage RST, the projection optical system pL, and the wafer stage WS). In a service room with a low degree of cleanliness that is separate from the installed clean room, the optical system for adjusting the optical axis called the beam matching unit is connected to the room η through at least a part including an optical system for transmitting light (not shown). . The light source 16 controls the on / off of the laser beam LB output and the i pulse of the laser beam LB according to the control information TS from the main control device 50. Frequency), center wavelength and spectrum half-value width (wavelength width). Describe the illumination optical system i 2 'with ·· beam shaping and uniform illumination 35 200400540 academic system 20 (including cylindrical lens, beam amplifier (ν σσ ^ not shown), and optical product benefit (homogenizer) 22, etc.) , Illumination system aperture diaphragm 24, ^ th relay lens 28A, 2nd relay lens 28B, fixed reticle curtain tender, movable reticle curtain 30B, optical path bending mirror M and condenser lens M Special. As the optical integrator, a fly-eye lens, a rod-type integrator (inner-surface reflection-type integrator), or a diffractive optical element can be used. For the sake of practicality, hereafter also referred to as the sorrowful system uses a compound eye lens as the optical integrator 22, the eye lens 22. The aforementioned beam shaping / illumination uniformity optical system 20 is provided through a light transmission window 17 in the room 11 and is connected to an optical system for light transmission (not shown). The beam shaping / illuminance uniformizing optical system 20 shapes, for example, a cross-sectional shape of a laser beam LB that is transmitted through a window 17 by a light source 16 that emits pulse light through a pulse using a cylindrical lens or a beam amplifier. Furthermore, in this beam shaping / illuminance uniformity optical system 20, the laser beam is passed through a coarse energy adjuster (not shown, equipped with an ND filter), and the transmittance can be changed in multiple stages or continuously in a proportional series. ), A plurality of diffractive optical elements in interchangeable configuration, movable cymbals (cone, polyhedron, etc.) along the optical axis of the illumination optical system, and an optical unit (not shown, which includes a zoom optical system) (At least one of them), reach the learner integrator 22 first. When the optical integrator 22 is a fly-eye lens, the optical unit can change the intensity distribution of the illumination light on the incident surface. When the optical integrator 22 is an internal reflection integrator, the illumination light can be changed on the incident surface. By changing the incident angle and the wide range, the light quantity distribution (secondary light source size and shape) of the illumination light on the pupil surface of the illumination optical system is changed, that is, the illumination condition of the reticle R is changed. In addition, the lighting 36 200400540 has tried to suppress the loss of light when changing the lighting conditions. In addition, the eye-opening 22 'located on the exit side of the beam shaping illuminance uniformity optical system 20 is used to illuminate the reticle R with a uniform illuminance distribution. The surface light source (2 = light source) composed of a plurality of point light sources (light source images) is formed on the exit-side focal plane that is arranged approximately one to one with the optical plane and the optical system. Hereinafter, the laser beam emitted from the secondary light source is referred to as a light EL ". In addition, a plurality of aperture apertures (for example, aperture apertures (general apertures) formed by general circular apertures) may be disposed near the focal plane on the exit side of the fly-eye lens 22 at approximately equal angular intervals; Aperture (small iris) used to reduce = coherence coefficient (σ) value formed by small circular openings; annular aperture diaphragms (annular diaphragm apertures) for annular lighting and deformed light source method , The illumination system opening aperture formed by a circular plate-shaped member such as a deformed opening aperture formed by eccentrically disposing a plurality of openings. In this case, the opening aperture plate of the illumination system is used at the same time as the aforementioned optical unit, and any opening aperture can be selectively set on the optical path of the illumination light EL, whereby the pupil surface of the illumination optical system can be performed. The light quantity distribution of the above illumination light (the size and shape of the secondary light source) can change the lighting conditions of the reticle R. In particular, even in the lighting condition which cannot be set using only the aforementioned optical unit, the opening aperture of the lighting system can be set, the amount of light loss can be reduced, and the lighting condition can be simply set. On the optical path of the illuminating light EL emitted from the fly-eye lens 22 (or the aperture diaphragm of the lighting system), a relay optical system (the first relay lens 37 200400540 28A and the second, ', and see 28B) Composition, the relay lens is between the fixed line sheet curtain 30A and the movable gong to the μ production quilt /, and the movable line sheet curtain 30B). The fixed graticule curtain 30A is a pattern surface from the conjugate surface to the graticule r, which is slightly defocused. The rectangular opening of the rectangular lighting area Mr on the regular graticule R is opened in the solid solid graticule curtain. Near 30A, the position corresponding to the scanning direction (γ car direction, H 1+ in the left and right direction on the paper in the picture) and the width-positioned / movable graticule with variable openings Curtain 30b, at the beginning and at the beginning of the beam exposure, is covered by a movable reticle | 30B, which further restricts the lighting area to prevent unnecessary exposure. X, the movable graticule curtain 3GB corresponds to the non-scanning direction orthogonal to the scanning direction (X axis: direction: the direction orthogonal to the paper surface in the i-th figure). The width of the opening is also variable. To transfer the pattern of the reticle R on the wafer, adjust the width in the non-scanning direction of the illumination area. -On the light private of the illumination = EL behind the second relay lens 28B constituting the relay optical system, a curved reflector M (for reflecting the illumination light E1 passing through the second relay lens 28B toward the reticle) R) A focusing lens 32 is arranged on the optical path of the illumination light el behind the curved mirror M. 々In the above configuration, the entrance surface of the fly-eye lens 22, the arrangement surface of the movable reticle curtain 30B, and the pattern surface of the reticle & are set to be optically conjugate, and the exit of the fly-eye lens 22 The light source surface (the pupil surface of the illumination optical system) formed by the side focus surface, and the Fourier transform surface (the exit pupil surface) of the projection optical system pL are set to be optically conjugated to form a Kohler illumination system . If the function of the lighting system constituted in this way is briefly explained, the laser beam LB emitted from 16 200400540 light source 16 pulses enters the beam-shaping uniformity optical system 20, the cross-sectional shape is shaped, etc., and then it is injected into the compound eye Lens 22 As a result, the aforementioned secondary light source is formed on the exit-side focal plane of the fly-eye lens 22. The illumination light EL emitted from the above secondary light source reaches the fixed reticle curtain 30A through the first relay lens 28A ′, and further passes through the opening of the fixed reticle curtain 30A and the movable reticle curtain 30B. And the second relay lens 28B, the optical path is bent vertically downward by the mirror μ, and then the condenser lens 32 is used to illuminate the reticle R held by the reticle stage rst with a uniform illuminance distribution. Rectangular lighting area IAR. A reticle R is mounted on the reticle stage RST, and is sucked and held by an electrostatic chuck (or vacuum chuck) or the like (not shown). The reticle stage rst is formed by a drive system (not shown) in a horizontal plane (χγ plane) that can be driven minutely (including rotation). The reticle stage RST is, for example, a reticle stage driving unit including a linear motor, which is not shown, and is perpendicular to the optical axis IX of the illumination system (which coincides with the optical axis Aχ of the projection optical system PL described later). In the plane, it can be driven minutely (including rotation around the z axis), and it can be driven at a scanning speed that specifies a predetermined scanning direction (here, the γ axis direction). The position in the XY plane of the reticle stage RST is set or formed on the reflective surface of the reticle stage RST by means of a reticle laser interferometer (hereinafter, referred to as a "reticle interferometer") 54R, for example, can be detected at any time with a resolution of about 0.05 to about. The position information from the reticle interferometer 54 on the other hand and the reticle stage RST is supplied to the main unit 40 200400540 which is set outside the body chamber u and is flat 50. The main control device 50 drives and controls the reticle stage RST 0 through the reticle stage driving unit (not shown) according to the position of the reticle stage rst. The material used for the grommet R must be added to the light source used. That is, when KrF excimer laser and ArF excimer laser are used as ^ this uses synthetic crystals such as synthetic quartz, fluorite, or fluorite-doped ^ but when using F2 laser, it must be fluorite, etc. It can be formed by crystallizing or fluorinated quartz. / The optical system PL has just been described, for example, using a telecentric reduction on both sides ^ 5 system. Qian Ying Optics' PL projection magnification is, for example, 1/4, 〆6, etc. Therefore, as described above, if the illumination area EL · is used to illuminate the moon area 1AR of the marking line, it will pass through the projection optical system. PL 'makes the circuit pattern of the reticle r in this lighting field: the small image formed on the crystal HW shared with the lighting field IAR ... the illumination field (exposure field) IA of EL. As for projection optics糸 # ΡΤ π ^ y 糸 、, PLPL · Er Lu, using a refractive system (only composed of a number of refractive optical elements (lens elements) 1 of the number of 1 1020 pieces. In the composition of the projection optics Multiple lens elements 13 of the system PL 13 lens elements 13 on the surface side (the reticle side) are for the sake of simplicity. Here are 5 lens elements 13, n, w _ performance correction. 4S 1 2 133,丨 35. The movable lens that can be driven from the outside is formed by the imaging elements n ~ 8 '. The lens is held in a lens holder that does not pass through a double structure (not shown), and 1 ~ 135 are held at Inner lens holder Two inner lens holders are driven by a drive element (not shown) For example, 200400540 piezoelectric elements, etc., support the outer lens holder with 3 points in the direction of gravity. And 'the driving voltage is independently adjusted for these driving elements, thereby forming 70 lenses 131 ~ 13S of each lens can be on the z axis Direction (the optical axis direction of the projection optical system PL) and the tilt direction of the χγ plane (that is, the rotation direction (0 X) around the χ axis and the rotation direction (6) y) around the Υ axis) The other lens 7L 13 is held by the ordinary lens holder and held in the lens barrel. It is not limited to the lens elements 13! To 135, and it can also be configured to drive an aberration correction plate (optical plate, which is arranged to correct the projection) Pupil of optical system pL

A lens near the surface or on the image surface side, or aberration of the projection optical system pL, especially a non-rotationally symmetric component). Moreover, the degree of freedom (movable direction) that can drive these optical elements is not limited to three, but may be one, two or more. —Taste A pupil opening σ aperture 15 capable of continuously changing the numerical aperture (NA) in the vicinity of the pupil plane of the PL optical system PL is provided. In the case of opening σ aperture 15 in Λ light, for example, a so-called rainbow aperture is used. The pupil aperture stop 15 is controlled by a main control device 50.

As ί Ming: To make Xiao KrF excimer laser light and ArF excimer laser light to EL, you can also use fluoride crystals such as Rongshi or other synthetic quartz doped with ytterbium to form a projection optical system. When using laser light, the material used in the projection optical system PL: the lens is made entirely of fluoride crystals such as fluorite or the aforementioned dopant: round W, through a wafer holder (not shown), and electrostatic adsorption (Second real adsorption) and the like are held on the aforementioned wafer stage WST. 41 200400540 The wafer stage WST is arranged below the projection optical system PL. The wafer stage driving unit (which is composed of a linear motor, a voice coil motor (VCM), etc.) can be located on the χγ plane. Inward direction and z-axis direction drive in 3 directions, the tilt direction to the X and Y planes (which is the direction of rotation around the X axis and Y

The rotation direction (0y)) around the shaft can also be driven slightly. The wafer WST can not only move the scanning direction (Y-axis direction), but also can make multiple exposure areas on the wafer w relatively move each exposure area IA for exposure, and the structure j can also be orthogonal to the scanning direction. Moving in the non-scanning direction (x-axis direction), thereby repeating the scanning exposure operation of each irradiation area on the wafer W, and the step scanning operation of moving to the acceleration start position for the next irradiation exposure . The positions in the XY plane of the wafer stage WST (including the rotation around the z-axis (0 Z) rotation) are set or formed on the reflecting surface of the wafer stage WST by a wafer laser interferometer (hereinafter, referred to as "Wafer interferometer") 54W "For example," the detection is performed at any time with a resolution of about 0.5 to 1 nm. The wafer interferometer 54W includes a plurality of multi-axis interferometers (with a plurality of length measuring axes). With these interferometers, the rotation of the wafer table WST (Θ z square yawing, Θ y rotation) can be measured by energy. (Pitching) (rotation around the Y axis), and θ x rotation (rotation around the X axis). Position information (or speed information) of the wafer table WST detected by the wafer interferometer 54W is supplied. To the main control device 50. The main control device 50 controls the position of the wafer stage WST through a wafer stage driving unit (not shown) based on the above-mentioned position information (or speed information) of the wafer stage WST. On the wafer stage WST, the alignment system ALG 42 which will be described later is formed. 200400540 Baseline measurement base & ^ + 屺 屺 寺 reference mark formed by the reference mark phase M, the surface of which is fixed approximately to the wafer W The surface is the same height. The 'Wait table W91T + n + + side (right side of the paper surface in Figure 1) of the royal face is a woman's dress. Wavefront aberration S measuring device 80. This wavefront aberration measuring device 80 is hollow as shown in FIG. 82, # party light optical system 84 first ^ ,, Mi (switchgrass is constituted by a plurality of optical elements, the light-based dry 7C member position relationship of both piece goods, arranged in the interior of the housing of the helmet μ),

And a light receiving unit 86 (arranged at one X-side end portion inside the housing 82).

The frame 82 is composed of a member (with a ζ-shaped cross-section in the shape of an L, forming a space inside), and at the uppermost end in the ㈤ direction, a circular opening 82a is formed in plan view (viewed from above) so that the frame 82 The light above 82 enters the internal space of the frame 82. A glass cover 88 is not provided to cover the opening 82a 'from the inner side of the housing 82. On the glass cover 88, a light-shielding film with a circular opening is formed in the center by steaming metal such as metal, and the light-shielding film is used to measure the wavefront aberration of the projection optical system pL. The surrounding unintended light hits the receiving optical system material. The light-optical optical system 84 is an objective lens 84a, a relay lens 84b, a 'cold mirror 84c, and a collimator lens 84d (which are sequentially arranged from top to bottom below the glass cover inside the housing 82). It is sequentially arranged on the X side of one of the curved mirrors 84 c) and a micro lens array 8 complement. The curved mirror 84c is 45 °. The inclination is set so that the optical path of the light incident on the objective lens 84a vertically downward from the curved mirror 84c is bent toward the collimator lens 8 牝 from above. In addition, each optical member constituting the light receiving optical system 84 is fixed on the inside of the frame 82f of 43 200400540 'through a holding member (not shown) and fixed. The plurality of small convex lenses (lens elements) of the aforementioned M mirror P 镜 84e are arranged in an array shape on a plane orthogonal to the optical path. The light-receiving unit is composed of a light-receiving element (consisting of a two-dimensional CCD or the like) and an electric circuit such as a charge transfer control circuit. The light receiving element enters the objective lens 84a and receives all light beams emitted from the microlens array 84e, so that it has a sufficient area. X. The measurement data of the light receiving unit 86 is transmitted to the main control device 50 through a signal line (not shown) or wireless transmission.

By using the wavefront aberration measuring device 80 described above, the measurement of the wavefront aberration of the projection optical system PL can be performed using the body (that is, the state where the projection optical system and the system PL are assembled in the exposure device). A method of measuring the wavefront aberration of the projection optical system PL using the wavefront aberration measuring device 80 will be described later.

In FIG. 1, an exposure device 100 of this embodiment is provided with a multi-focus position detection system (hereinafter referred to simply as a "focus detection system") of an injection method. The focus detection system is composed of an illumination system 60a (having a light source controlled by the main control device 50 to be turned on / off, and irradiated from an oblique direction to the optical axis AX toward the imaging surface of the projection optical system to form a plurality of pinholes or The imaging beam of the slot image), and the light receiving system 60b (the reflected beam on the surface of the wafer receiving these imaging beams). In addition, the detailed configuration of a multi-point focus position detection system similar to the focus position detection system (60a, 60b) of this embodiment is disclosed in, for example, Japanese Patent Application Laid-Open No. 6_2834403 and corresponding US Patent No. No. 5,448,332, etc., the present disclosure cited the disclosures of the above-mentioned bulletin and the US patent as part of the description of this specification. 44 200400540 In addition, many points described in the above-mentioned publication and U.S. patents are particularly recognized as private point focus position detection systems, which not only detect the position of the wafer w1 (which is in the field of exposure) The number of points in the IA and y in the non-scanning direction is set to the number of points p and the number of parallel lines, which are respectively parallel to the projection light Γ X (z-axis direction), and has the ability to first read the wafer w in # 彳 田 direction. It can be, and can be, also, it is "2" but even if it does not have these functions parallelogram or other shapes. ~ The shape of the beam before the shot can also be the main control device 50, which is from the light receiving system 60b during scanning exposure, etc. The focus shift signal (Scatter II, (1 ,,, ^ Tiger)), for example, according to the S-curve signal II: i: product: move: (not shown), to control the position of the wafer ... and the χγ plane The main control device 50 is used to perform the auto-focusing and the rafting force. The main control device 50 measures (60a, _) of the wavefront aberration described below to perform θ Focus position detection system ... Η Measuring device ... Slope measurement of position measurement. Wavefront aberration measurement device 80 = exposure device is equipped with an off-axis alignment system. With off-axis (off · call square ^ fresh "alg, the wafer held on the WST w day day w mouth The position of the reference mark on the board FM + Jiwei Baoli, then Temple. Regarding the alignment system ALG and: 'for example, the detection mark of a wide band that will make the photoresist on the wafer insensitive to the light band, Taking the reflected light from the ^ ^ ^ Ding 5 as a result, use a shirt (and so on) to photograph the image of the object mark imaged on the light-receiving surface. The more unillustrated index image 'Use image processing to output these photographic signals. Equation 45 200400540 A sensor of the FIA (Field Image Alignment) system. It is not limited to the FIA system. Of course, it is possible to use the alignment sense alone or in an appropriate combination to cry. It is a method of irradiating a coherence detection light on an object. The symbol detects the scattered light or the diffracted light generated from the target symbol, and interferes and detects two diffracted lights (for example, the same order number) generated from the target symbol. The exposure device 100 of this embodiment is omitted, though The figure is located above the reticle R. A pair of reticle alignment detection system, which consists of a TTR (Through Ugh The Reticle) alignment system, which uses the exposure wavelength (through the projection optical system PL to observe the reticle at the same time) The reticle mark on R and the corresponding reference mark on the reference mark plate), for these reticle alignment detection systems, for example, ^ and: JP + 7-176468 and corresponding Gazette No. 5,646,413 and other similar constituents. To the extent permitted by the domestic laws of the designated country or selected country designated by the applicant ', this case cited the above-mentioned bulletin and the US Patent " As part of this book. ㈢ 心 秋 之 In the second picture, the aforementioned control system is mainly composed of the aforementioned master device M. The main control device 50 is composed of work, station (also called microcomputer, and pots are composed of coffee, job, editor, etc.), etc. In addition to performing various control actions of ㈣, it also controls the entire, for example, W Wave home. The main control device 50 performs the exposure operation in a consistent manner, for example, the step between irradiations of the WST, the exposure timing, and the like. The wafer stage is locked, and the main control device 50, for example, 'the connection is composed of hard disks', the input device L, the clothing 42 (the system., The system is directed by the keyboard, the air, etc.) 46 200400540: Composition) and display devices such as CRT monitors (or liquid crystal displays). The flat control unit 50 is connected via a communication network such as a LAN to work with a simulation computer 46 such as a personal computer. In the simulation computer 46, the imaging simulation software (i.e., the 'imaging simulator') set with the optical model of 1 GG is set to ^ 7. = Times' Explains the wavefront image of the exposure equipment during maintenance and so on. In the following description, in order to simplify the aberration of the light-receiving optical system 84 in the wavefront aberration measuring device 8G, it is considered to be small enough to be ignored. During normal exposure, the wavefront aberration measuring device 80 is removed from the wafer table. When measuring the wavefront, "first, it is called" by the engineer ", etc.), the wafer is measured. Taiwan WST :: Stop: Order to install wavefront aberration measuring device 8 ... industry. When the wave surface in-line measuring device 80 is installed, when measuring the wave surface, the wave surface aberration measurement event ^ 80 is fixed to the predetermined reference surface through screw inspection or magnets to cover the movement stroke of the wafer table operation. (This refers to the + side surface. After the above installation is completed, the operator will respond to start the measurement: the input of the command, the main control device 50, and its wavefront aberration measurement device 8: Located below the alignment system and system ALG 'is to move the wafer table WST through the wafer table drive (not shown). In addition, the main control device 50 detects an alignment mark (not shown) provided in the wavefront aberration measurement device 80 through the alignment system ALG, and the amount of the wafer interferometer M: at this time is measured based on the detection result. The measured value 'calculates the position coordinates of the alignment marks, and obtains the correct position of the wavefront aberration measurement device 80. And "after measuring the position of the wavefront aberration measuring device ⑽ 47 200400540", the main control device 50 performs the wavefront aberration measurement as described below. First, the main control device 50 carries a non-illustrated measurement marking sheet (hereinafter, referred to as a "pinhole marking sheet") formed by a pinhole pattern through a not-shown marking sheet carrier. On the reticle stage RST. The pinhole reticle is a reticle that forms pinholes at a plurality of points on the pattern surface (roughly becoming an ideal point light source and generating pinholes of spherical waves). In addition, for example, in order to coincide with the optical axis Ax of the projection optical system PL, when setting the pinhole reticle, a plurality of pinholes are arranged in the lighting field IAR, and the projection images are formed in the projection optical system respectively. In the field of view of the PL, a plurality of points of wavefront aberrations are measured (the first to nth measurement points described later). In the pinhole reticle used here, a diffusing surface is provided thereon, and the pupil surface of the projection optical system PL is substantially comprehensive, so that the light from the pinhole pattern is distributed, and the projection optical system is thereby used. The comprehensive pupil surface of PL is used to measure wavefront aberration. In this embodiment, since the aperture diaphragm 15 is provided near the pupil surface of the projection > optical system PL, the wave surface aberration is measured at the pupil surface defined by the aperture diaphragm 15. _ After carrying the pinhole reticule, the main control device 50 uses the aforementioned reticule alignment detection system to detect the reticule alignment mark formed on the pinhole reticule. 'According to the test result, the pinhole marker The thread is aligned to a predetermined position. Thus, the center of the pinhole reticle is approximately the same as the optical axis of the projection optical system PL. Then, the main control device 50 supplies the control information TS to the light source 16 to cause the laser beam LB to emit light. As a result, the illumination 48 200400540 from the illumination optical system 2 is irradiated on the pinhole reticle. In addition, light emitted from a plurality of pinholes of the pinhole reticle passes through the projection optical system PL, is condensed on the image plane, and the pinhole image is formed on the image plane. Secondly, 'the imaging point of the main control device 50 in the image of any pinhole (hereinafter referred to as the eyelet pinhole) on the pinhole reticle is to coincide with the approximate center of the opening 82a of the wavefront aberration measurement device 80' While monitoring the measured value of the wafer interferometer 54W, the wafer stage WST is moved through the wafer stage driving section (not shown). At this time, according to the detection result of the focus position detection system (60a, 60b), the main control device 50 places the top of the glass cover 88 of the wavefront aberration measuring device 80 on the image surface of the pinhole image, and transmits it through The wafer L driving unit (not shown) drives the wafer table operation slightly in the two-axis direction. At this time, the tilt angle of the wafer table WST is also adjusted as necessary. As a result, the image beam of the eyelet pinhole passed through the central opening of the blood glass 88 of the glass cover 88, and entered the optical optical system 8 4. The 86 light receiving element receives light. Specifically, a spherical wave is generated from a pinhole on the eye of a pinhole reticle, and the spherical wave is transmitted through a projection optical system PL and an objective lens (which constitutes a wavefront aberration measurement device 80). The optical system for receiving light is shown in figure 84), relay lens 84b, reflector 84c, and collimator lens 84d, which form a powerful beam and irradiates the microlens array 84e. In this way, projection optics can be used Subsequent to the microlens array 84e, it is divided. Moreover, by the lens elements of the microlens array 84e, the respective light is focused on the light-receiving surface of the camera and the light-receiving part, respectively. The surface forms a pinhole image.

At this time, if the "right projection well study DT" L is the ideal optics 49 200400540 system without wave surface aberrations, the wave surface of the projection optical system PL $ pupil surface becomes the wave surface of the rational phase (here is a plane). As a result, The flat Γ beam incident on the microlens array 84e becomes a plane wave, and the wave surface is an ideal wave surface. In this case, as shown in Fig. 3A, at the position on the optical axis of each lens element constituting the microlens array, an imaging point is formed. Image (hereinafter, also referred to as "dot"). However, in the projection optical system PL t, generally, because the wavefront aberration exists, the wavefront of the parallel beam incident on the microlens array 84e is shifted from the rational phase wavefront according to the offset (that is, the wave faces the ideal wavefront) After the completion of each point as shown in FIG. 3B, the position is shifted from the position on the optical axis of each element of the lens array 84e. In this case, the position shift from the position on the optical axis of each lens element) of each point is the oblique production corresponding to the wave front: Also, the light beams incident on the light-concentrating points on the light-receiving element constituting the light-receiving section 86 Occupying the image beam) 'is photoelectrically converted by the light-optical element, and the photoelectric conversion signal is transmitted to the main control device 150 through the electric circuit. The main control setup 50 calculates the imaging position of each point according to the photoelectric conversion signal, and =: the calculation result and the known reference point position data are used to calculate the position offset (△, "), and then stored in the RAM. At this time In the main control device, the measured value of the wafer interferometer 54w at this time is%, ^). As described above, if the wavefront aberration measurement device of the imaging point of the eyelet pinhole image is used, i 8〇, When the measurement point image position shift is completed, the main control device 50 is located at the imaging point of the bottom-pinhole image, and the wafer table WST is moved so that the approximate center of the opening 82a of the plane aberration measurement device 80 is consistent. After this movement, the beam is emitted, the laser beam LB is emitted from the light source 16 by the main control device 50 in the same manner as above, and the imaging position of each point is calculated by the main control device * 50 200400540. Then, the same measurement is sequentially performed at the imaging points of other pinhole images. In this way, at the completion of the necessary measurement stage, in the RAM towel of the main control device 50, the aforementioned imaging of each pinhole image is stored. Point offset

(△ f, △ and the coordinate data of each imaging point (which is the measurement value of the wafer interferometer 54W (际 丨, Υ) when measuring the imaging point of each pinhole image). Also, in During the above measurement, the illumination light EL㈣ can also be irradiated to all pinholes, or a movable graticule can be used to cover 30B. For example, the position and size of the lighting area on the graticule can be changed in each pinhole Use illumination iris to illuminate only the pinhole or part of the eye (at least including the pinhole of the eye) on the reticle.

Secondly, the main control device 50 corresponds to the imaging point of the pinhole image according to the principle explained below according to the positional offset data of the imaging point of each pinhole image, △ ^) and the coordinate data of each imaging point stored in the RAM. That is, the wavefront (wavefront aberration) from the measurement point (evaluation point) to the ηth measurement point (evaluation point) in the field of view of the projection optical system PL. Here, the calculation will be described later according to a conversion program. The coefficients of the terms (Chagneck term) of the stripe Chase polynomial of the formula (3) (hereinafter referred to as the Chagneck polynomial ") (for example, 'the first term is Lou Wen Z, the coefficient of the 37th term ζ37 ). In the present embodiment, the stripe Chaney polynomial is used as the Chase polynomial, and the following description will be made. In this embodiment, the calculation is performed according to the above-mentioned position shift (Δg, △ 々) according to the conversion program, thereby obtaining the wavefront of the projection optical system pL, that is, the position shift (ΔP), which still reflects the ideal wave surface. 51 200400540 The slope value of the wavefront, on the contrary, the wavefront can be restored based on the positional offset (, —,, f, △). In addition, the above-mentioned positional offset (△ $, ^ Q々) and the physical relationship with the wavefront It can be known from the sexual relationship that the wavefront of this embodiment is calculated, j: y 1 糸

Shack-Hartmann wave front calculation principle. Next, a method for calculating the wavefront based on the above-mentioned positional offset will be briefly described. '' As mentioned above 'The position shift (△ f, △) is the oblique production of the corresponding wave surface. The position shift (△ $, △ π) is integrated, and the shape of the funnel surface (strict: "from the reference surface (ideal wave surface) The offset of). Let the formula of the wave surface (which is the offset wave surface from the reference surface) be w (x, y) and the proportionality coefficient be the relationship of the formula (υ, (2). The relationship holds. Dx… ⑴ Αη = ^ dy… (2) because it is still not easy to integrate, it only represents the strong point of your slave s ^, and the slope of the leather surface 'so the surface shape Li field is adapted to this formula. In this case, the series system chooses the orthogonal two t t polynomial The series is suitable for the expansion of the axisymmetric plane, and the circumferential direction is the expansion of the diagonal series. ^. W If the wavefront W is represented by the polar coordinate system (ρ, θ) on the right, it can be expanded as shown in the following formula (3) Ν (Αθ) ... ⑶

Since it is the patrilineal system of the father, the value of A can be used to define the coefficients Zi and i of each pair. Perform some filtering with the appropriate 7 monks. In addition, for example, if the number zi is instantiated at the same time from the first term to 筮 π T5 right /, coefficient, and fi (ρ, 0) (taking ρ as an independent variable 52 200400540 direction diameter polynomial), the following list Work shown. However, the 37th item in Table i is equivalent to the 49th item in the actual Zernike polynomial, but in this specification, the item 37 (item 37) is used. That is, in this case, the number of polynomial terms is not particularly limited.

z2 Z3 Z4 Z5 Z6 Z7 Z8 Z9 zzzzzzzzz pcosQ p sin0 2ρ2Λ p 2cos2 0 p 2sin2 Θ (3 p 3-2 p) cos Θ (3 p 3-2 p) sin Θ 6p4-6p2 + lp 3cos3 Θ p 3sin3 0 ( 4 p 4-3 p 2) cos 0 (4p4-3p2) sm0 (10 p 5-12 p 3 + 3 p) cos Θ (10 p 5-12 p 3 + 3 p) sin Θ 20/0 ^ 30 / 9 ^ 12 ^^ 1 p 4cos4 Θ p 4sin4 Θ Z19 ^ 20 Z21 ^ 22 ^ 23 ^ 24 z25 Z26 Z27 ^ 28 Z29 Z30 Zsi Z32 Z33 Z34 Z35 Z36 Z37 (5p5-4p3) cos3 0 (5 p 5-4 p 3) sin3 Θ (15p6-2〇p4 + 6p2) cos0 (15p6-2〇p4 + 6p2) sin0 (35 p 7-60 p 5 + 30 p 3-4 p) cos 0 (35 p 7-60 p 5 +30 p 3-4 p) sin 0 70psA40p6 + 90p4-20p2 + lp 5cos5 0 p 5sin5 Θ (6 p 6-5 p 4) cos4 Θ (6 p 6-5 p 4) sin4 Θ (21 p 7-30 p 5 + l〇p 3) cos3 0 (21 p 7-30 p 5 + l〇p 3) sin3 Θ (56 p 8-105 p 6 + 60 p 4-10 p 2) cos2 Θ (56 p 8- 105 p 6 + 60 p 4-10 p 2) sin2 Θ (126 p 9-280 p 7 + 210 p 5-60 p 5 + 5 p) cos (9 (126 p 9-280 p 7 + 210 p 5- 60 p 5 + 5 p) sin Θ 252 p 10-630 p 8 + 560 p 6-210 p 4 + 30 p 2-l 924 0 l2-2772 p 10 + 3150 p 8-1680 p 6 + 420 p 4- 42 p 2 + l

Actually, this differential system is detected by the above-mentioned position shift, so the adjustment must be performed for the differential coefficient. The polar coordinate system (X = P COS 0, y = p) is expressed by the following formulas (4) and (5). 3W _ dx dW 1 dW. A = cosy --- sm0 dp p 8Θ… (4) dW dy " dW. Ω, ^ dW n = -sm ^ +-cos (9 dp p d0… ⑶ Inzanne The differential fractal of the grams polynomial is not an orthogonal system, so adjustment must be performed on 53 200400540 square Λ. Because the information (offset) of a point image imaging point is in the direction of the PL field of view, the pinhole number is Corresponding to the projection optical system, for example, the number of measuring points (§ parity points), in this embodiment, to simplify the number of programs ==), the observation method represented by the above formulas ⑴ ~ ⑺ (i 小 查 之 = The terms of the crew type correspond to optical aberrations. In addition, the low-order terms a "~, and Seidei aberrations correspond approximately. Using Chase, it is possible to find the wavefront of the projection optical system PL. Aberrations: The principle described by the conversion ΓΓ determines the operation steps of the conversion program, and according to the conversion operation of the private formula,

The second output in the field of view corresponds to the wavefront information (wavefront aberration) of the measurement point to the nth measurement point in the projection optical system PL. Here, the vignette coupon $) can be obtained. The coefficient is, for example, a coefficient 217 from the first term to a coefficient 237 from the 37th term. The database of the wavefront aberration change table of the projection optical system PL is stored in the memory device 42. Here, the so-called wave surface aberration change table refers to the change table formed by the shell material group. The data group arranges the data according to a predetermined rule. The data indicates the change of the unit adjustment amount of the adjustment parameter (pot system = Equivalent to the equivalent model on the projection optical system PL, perform the simulation 'to obtain the simulation result, and optimize the formation of the wafer on the wafer image, and correspondingly correspond to the projection optical system pL The relationship between the imaging performance of a plurality of measurement points in the field of view (specifically, the data of its wavefront, for example, the coefficients of the i-th to the 37th coefficients of the polynomial polynomial). In this embodiment, the driving scenes of the movable lenses 13 丨, " 2, 133, 134'135 are driven by the degree direction (driving direction) ~ mile ~ χ, β y], z 54 200400540, 0y5), said A total of 19 parameters of the circle w surface (wafer stage w: momentum, x, y) and Zhao Mingguang ELj driven by the degree direction ~ Ming Xian E] L wavelength offset △] as the above Adjustment parameters. Here, the steps for making the database of the wavefront aberration change table mentioned above are explained earlier. First, when installing specific optical software ''

Optical conditions of 'input exposure device' in the brain (for example, design values of projection optics ^ PL (numerical aperture NA, and various lens data, such as coherent α values (illumination σ)) or numerical aperture na of illumination optical systems, and Zhao EL Wavelength (exposure wavelength) and so on). Second, in the simulation computer village, enter any number of measurement points in the projection optical system PL field of view.

Next, input the data of each unit amount of the moving direction of the movable lens 13l to 13s (movable side :), the above-mentioned degrees of freedom of the ® WS surface, and the deviation of the wavelength of the illumination light. For example, if an instruction to drive only 13l units of the movable lens is input in the + direction shifted in the z direction, the simulation computer 46 is used to calculate a predetermined 9 1 = measuring point in the field of view of the projection optical system P1. The data of the change amount of the ideal wave surface of the first wave surface, such as the coefficient change amount of each item of the Zernike polynomial (for example, the first item to the 37th item). The data of the change amount is displayed on the simulation computer. On the screen of the 46 display, and the change amount is recalled in the memory with the parameter .PARAipi. ° Secondly, if it is tilted in the Y direction (+ direction of rotation around the X axis, if the command to drive only the movable lens 13! Unit amount is input, then only the second measurement computer 46 is used to calculate the first measurement point for the first measurement point. 2 wave surface data, ^ 55 200400540 T, calculate the change amount of the above-mentioned coefficients of the Chase grams polynomial, the change amount data is displayed on the day surface of the above display, and the change amount is stored in the parameter PARA2P1 In memory

Next, if the command to drive only 13l units of the movable lens is input by tilting in the X direction (rotation of 0 around the y axis + magic direction), the 3rd measurement point is calculated by the brain 46 for simulation, and the 3rd is calculated. Wave surface data, for example, ^ 'Calculate the variation of the above-mentioned coefficients of the Chase polynomial. The data of the variation is displayed on the screen of the above display, and the variation is stored in the memory with the parameter PARA3P 1 After that, using the same steps as above, the input of each measurement point from the second measurement point to the second measurement point is performed, and the z-direction shift, the y-direction tilt, and the x-direction of the movable lens η are performed each time. When the direction tilt command is input, the computer 46 of the simulation is used to calculate the i-th wave surface and the second wave-th wave surface of each measurement point. For example, the data of each change amount is displayed as 'significant.' In writing, count g-up and record PARA1P2, PARA2P2, ARA3P2 ......... PARA1Pn, PARA2Pn, PARA3Pn

In memory. " ㈣For other m 13 ^ 4, 135, also use the same + party A ^ 仃 each! The input of the measuring point and the commander who only drive a unit amount in the direction of the respective degree, respectively, corresponding to this, by simulating electricity for driving only the movable lens U in the direction of the respective degree ;: 13 The first to nth measurement points of the world, calculate the wavefront data, all the changes in the Zernike polynomial, parameters (PARA4P1 RA5P1, PARA6pi, ... PARA15P1), parameter 56 200400540 ( PARA4P2, PARA5P2, PARA6P2, ..., PARA15P2), parameters (PARA4Pn, PARA5Pn, PARA6Pn, ..., PARA15Pn) are stored in the memory. For wafer W, the same steps as described above are used to input each measurement point, and to input only a unit amount command in the direction of each degree and in the + direction, respectively. Corresponding to this, a simulation computer is used. 46. Calculate wavefront data for the 1st to ηth measurement points when only the wafer W unit quantity is driven in the directions of Z, 0 X, and 0 y degrees of freedom, for example, calculate the various changes in the Zernike polynomial. The parameters (PARA16P1, PARA17P1, PARA18P1), parameters (PARA16P2, PARA17P2, PARA18P2), ... parameters (PARA16Pn, PARA17Pn, PARA18Pn) are stored in the memory. Further, the wavelength shift is also the same as the above In the same steps, input of each measurement point and the instruction input of shifting the wavelength by only a unit amount in the + direction are corresponding. Accordingly, the simulation computer 46 is used to shift the wavelength by only a unit amount in the + direction. For the first to nth measurement points, calculate the wavefront data. For example, calculate the changes of the various Zernike polynomials. The parameters (PARA19P1, PARA19P2 ........ PARA 1 9Pη) are stored in the memory. Here, the above parameters PARAiPj (i = 1 to 19, j = 1 to η) is a column matrix (vertical amount) of 37 rows and 1 column. That is, if η = 33, the adjustment parameter PARA1 becomes the following formula (6). Also, The parameters PARAiPj are all matrices. As for the following formula (6), for the sake of convenience, a row matrix-like expression is used. 57 200400540 PARA \ PI = [ZU Zl2 …… 2137], PARA \ P2 = [Z2 > 1 Z2f2 ·····. Z2 1: JU (6) PARAlPn = [z33 j Z33 2 ...... 233 37 j Moreover, regarding the adjustment parameter PARA2, it becomes the following formula (?). Λ4Question 1 = [ZU Zu… ··· 2137]]

PARA2P2 = [Z2fl Z2 > 2 ····· 223J: J… ⑺ PARAlPn = [z331 Z33 2 ...... 233 37 j

Similarly, the other adjustment parameters PARA3 to PARA19 also have the following formula (8). PARA3P1 = 1 ^ 1,1 ^ 1,2 r · Α37] PARA3P2 = lU2,2 · • · Α37] PARA3Pn = [^ 33,1 ^ 33,2 · • ... 233,37] PARA19P1: = 1¾¾… r PARA19P2 No. 2,37] PARA \ 9Pn:-[Z331 ZZ33 2 · ^ 33,37.

Therefore, in this way, the column matrix (longitudinal quantity, which is composed of the internal memory. The amount of change in the coefficients of the recalled Zernike polynomials) PARA 丨 p ~ PARA19Pn is concentrated on the adjustment parameters, Alternate side by side is used as a wavefront aberration change table of 19 adjustment parameters. As a result, a matrix (row and column) represented by the following equation (9) with the column matrix (vertical amount) PARA1P1 to PARA19Pn as elements is prepared. In Equation (9), m = 19. 58 200400540 PARA1PI PARA2P1 PARA1P2 PARA2P2 PARAmPl PARAmP2 • ⑼ PARAmPn PARAlPn PARAlPn Therefore, the lean material library (which is a projection optical system made by

The wave surface aberration change table) is stored in the memory device PL < Next, the method of setting the aforementioned 19 adjustment parameters of the movable lens Ui-Uj and the like used to adjust the imaging state of the image of the projection optical system PL in the exposure system 100 of the present embodiment is used to adjust the imaging state of the projection optical system PL. That is, the general adjustment method of the projection optical system PL includes a detailed description of its principle. "First, using the steps described above, a wavefront aberration measurement device is used to measure the wavefront aberration of the projection optical system PL. The measurement result is to find the corresponding measurement point (evaluation in the field of view of the projection optical system PL) Point) to the n-th measurement point (evaluation point) of the wavefront (wavefront aberration), that is, to find the terms of the Zernike polynomial ', for example, to find the coefficient of the i-th term & the coefficient of the 37th term Z37 is stored in the RAM of the main control device 50, etc. In the following months, the corresponding moment from the first measurement point to the nth amount can be expressed by the moment P car Q of the following formula (10). Wavefront (wavefront aberration) data of the measurement points.

Q • (10) 59 200400540 Also in the above formula (10), the elements of the early Q's Q, P! ~ Pη series of matrices (longitudinal unitary, which are respectively represented by 杳 Nixiuduo 笪, 圼 Chongxi terms Coefficients (Zi ~ Z37) of the 1st to 37th terms). t 'by the main device 5G', as described below, calculates the adjustment amount of the respective directions of the movable lens 131 to 3ΐ i / 5 and the adjustment amount of the respective directions of the wafer W. ..., the amount of wavelength shift of bright light EL. That is, the wavefront (wavefront aberration) at the measurement point corresponding to the i-th elimination point ~ # ^ times,, "", the lean material Q, the aforementioned database 丨 beam ten rhino (matrix 0), and the aforementioned 19 adjustments The relationship of the following formula (π) between the quantities P holds: n (11)

Q = 0 · P In the above formula (11), P is a column matrix (ie, a longitudinal quantity) composed of the elements (i.e., 19) represented by the following formula (12). ADJ1 ADJ2 P, • (12) ADJm

Therefore, 'the following M 1 is performed according to the above formula (12), and the calculation of the following formula (13) is performed, that is, by using the least square method, the elements Am of p can be obtained! I Λ receives I ADJ1 ~ ADJm, that is, Calculate the adjustment amount of the movable lens ^ ~~ 135 in the direction of the direction of the lens (target adjustment amount, the amount of the wafer W in each direction of the direction ^^, and the same adjustment 1 (target adjustment amount ), And the wavelength shift amount (target shift amount) of the illumination light EL. P = (Οτ · Ο) 1 · 〇τ · ο · / V ... (13) In the above formula (1 3) In 〇τ, the transpose matrix of your queue 〇, (〇τ · 〇) -1 is the inverse matrix of 200400540 (〇τ · 0). Therefore, the main control device 50 is a database in the memory device 42 Read into RAM in order, and calculate the adjustment amount ADJ1 ~ ADJm. Second, the main control device 50 supplies the command values of 13 ° ~ 1335, which drive the movable lens 13 ° ~ 35, respectively, according to the ADJ1 ~ ADJ15 stored in the memory device 42. The performance correction controller 48. Therefore, by the imaging performance correction controller 48, each of the driving elements that drive the movable lens 11 to 135 in the degree direction is controlled. At least one of the positions and methods of the movable lens Πι to ΐ35 is adjusted at the same time based on the applied voltage, and the 50 is used in the actual scanning exposure at z,. At the same time, the main control devices ^ X, Θ y are freely oriented.

The command value for driving the wafer w is supplied to the wafer stage driving section (not shown) to drive the wafer stage WST, so that in the exposure area IA, the wafer w is always maintained with the adjustment amount ADJ16 ~ ADJ18 is an equivalent way to adjust the value. In addition, at the same time as the above, the main control unit S 50 supplies a command to the light source 16 according to the adjustment amount ADJ19 to shift the wavelength of the lighting A EL. As a result, the optical characteristics of the projection optical system PL, such as variability, curvature of field, coma aberration, spherical aberration, and astigmatic aberration, are corrected. In addition, with regard to coma aberration, spherical aberration, and astigmatic aberration, both low-order and high-order aberrations can be corrected.

Next, an adjustment method for the projection optical system PL (which is for the purpose of adjusting the line width difference between the line pattern images in the orthogonal 2 axis direction performed by the exposure device 100 of this embodiment) is shown in FIG. 4 The flowchart is explained with reference to other drawings as appropriate. First, in step 102 of FIG. 4, the foregoing steps are used to measure the projection optical system difference using 61 200400540 wavefront aberration measuring device 80. The result is just the result, that is, the image of the projection optical system is obtained. 'η%) of the measurement point (evaluation point) of the check polygram for a long time, for example, to find hope] Miao Xi always go, the σ term " the coefficient of the 1st term of the Watt brother ~ 37 of 37 Zhi z37 'is then memorized in memory such as ram of the main control device 50. 2. In step 104, 'the following reticle for measurement described below: occasionally called "reticle Rt") is carried on the reticle # RST, two! W =: wafer (for convenience, abbreviated "Wafer ^")

mouth . The carrying of the wire segment Rt and the wafer WT are carried out at the position of the main frame display 50 or less' by a reticle-bearing circular carrier (not shown). 3

Here, the reticle Rt will be described with reference to FIG. 5. Fig. 5 is a plan view of the reticle & viewed from the pattern surface side. As shown in FIG. 5, the reticle RT forms a rectangular pattern ㈣pA, which is composed of a square: glass substrate, which is approximately the same as the lighting area IAR surrounded by a light shielding band SB in the central portion of the pattern surface. Its shape. A total of 33 measurement patterns for measurement patterns ΜΡι ～ Μρ ^ are formed inside the pattern area pA. The pattern for each measurement] MP'l ~ 33). For example, when the center of the reticle ruler (pattern area PA) and the optical axis AX of the projection optical system and the system pL are all set to this position, then The operator has placed the measurement points (evaluation points) within the effective field of view of the projection optical system PL for measuring the aforementioned wavefront aberrations. Each measurement pattern MP is a line width on the design including the line width (for example, 600 nm) in the design extending in the γ-axis direction and a line width in the design extending on the χ-axis direction as shown in FIG. 5. (For example, 600 nm) a second line pattern. If the projection magnification of 62 200400540 projection optical system PL is set to 1/4, and the first line pattern and the second line pattern are transferred to the wafer, there will be no spherical aberration or image on the projection optical system ρ. Under ideal conditions of various aberrations such as astigmatism, a line pattern image with a line width of 150 nm can be obtained for the images of the first line pattern and the second line pattern, respectively.

Also, on both outer sides of the pattern area p A on the X axis passing through the center of the pattern area PA (which coincides with the center of the reticle), a reticle alignment mark (M i, M2) is formed. The reticle Rt is a state where the reticle Rt is carried on the reticle table RST, and the pattern surface (the surface on the front side of the paper in FIG. 5) becomes the surface facing the projection optical system 卩 B. In Figure 5, the reticle alignment is performed in the next step 106. This reticle alignment is disclosed, for example, in Japanese Patent Application Laid-Open No. 7_176468 and the corresponding U.S. Patent No. 5,646,413. For detailed disclosure, the main control device 50 uses the aforementioned reticle alignment detection system to detect each The position deviation of the reticle alignment marks (M1, M2) formed by the reticle RT and the base ^ marks on the reference mark plate FM formed on the wafer table wst is adjusted according to the δH detection result. The position (including rotation) in the XY plane of the graticule table is performed to minimize the position deviation between the two. By the alignment of the reticle, the center of the reticle Rt is approximately the same as the optical axis of the projection optical system PL. In the next step 108, under the predetermined lighting conditions, each measurement pattern Mρ of the reticle Rt arranged in the lighting field IAR is used to pass the reticle stage RST and the wafer stage through the projection optical system PL '. The wst is statically transferred on the wafer wT, and the image of the measurement pattern mPj in the pattern area PA 63 200400540 (latent image) is a photoresist layer coated on the surface of the wafer w # τ. In addition, the main batch manufacturing device 50 is based on the detection results of the focus detection system β, and the bismuth measurement system (60a, 60b). Let the wafer WT surface pass through the wafer stage driving section (not PJ such as (not shown), and drive the wafer stage slightly in the z-axis direction. If necessary, the wafer △ is the tilt angle of the W stage WST It is also adjusted to move the wafer table WST step-by-step to the production gate. Of course, in multiple fields on the Japanese yen, wT, of course, it can also be transferred in sequence from £ 卩 & Pattern area PA. In the next step 110, μ, +, Buddha Μ υ Measurement chart on the above-mentioned private line piece RT

The wafer transferred by MPj is selected from the wafer table WST according to the instructions of the autonomous control device 50, and is placed in the exposure # ^ exposure to 100 inline (ilMin The connected unillustrated photoresist coating and developing device (_r devei.) Is called, and is conveyed by the unshown conveying system. In the next / step 112, the main control device 50 will supply instructions to the cold cloth developing device. The control system 'according to this instruction, by the unillustrated coating: developing device' on the crystal Wt, forming a photoresistance image of the measurement ㈣MPj is next-step 114 'The developed wafer ^ is the same as previously Similarly, it is carried on the wafer table WST. In the next step 116, the line width measurement of the photoresist of the measurement pattern MP on the wafer Wt is performed. The line width measurement is, for example, performed by the main The control device 50 moves the wafer stage WST in the XY plane, and uses the alignment system alg to sequentially photograph the photoresist images of at least i measurement patterns MPj on the crystal m Wt. Signal to perform predetermined processing (including calculations). As a result, the evaluation of each projection optical system pL 64 2004 00540 Price point (measurement 1

Mpj, find the line width L1 of the first line pattern 1 and the line width L2 of the second line pattern 2 and store them at the station), gp, the pattern width of each measurement pattern image (in this case, the photoresist image) The width of the image (such as the moon-shaped image refers to the photoresistance image) of the RAM and other devices in the body. 50 is based on the width L2 obtained from the above, and the relevant quantities are obtained and stored in the RAM and the like below. In step 11-8, the line width L1 of the measurement pattern MPj and the line measurement pattern MP ′ of the main control device are measured. The width difference △ L = u-L2 in the memory.

Here, 'for the plurality of measurement patterns MP and silk on the crystal K Wt e? Tian Jinde said the above fields, to P] the formation of resistance' is for each complex number-仃 the above line width measurement, line width difference Figure it out. This kind of feeling can also be obtained from the plural areas " to the line difference (for example, the simple average) of the pattern MP 丨 used as a measure for each measurement.妁 妁 — 里 说说 ”The green sees the critical L, and the effect of Qianjun reduces the measurement error of $, which can be higher; find out the measurement patterns Mp / ^ ^ ^ ^ Brother 1 line pattern (vertical, 彳

, ^ "Difference from the image line width of the second line pattern (vertical line pattern) (the line width difference of the vertical and horizontal lines")). In the next step 120, the value (size) of the coefficient of the twelfth term of the ^ term formula is based on the values of the 1 and 2 points of the branch line at the points AL, 里, and W, and the 9th term of a Coefficient Z9 〇 Regarding the amount of wavefront aberration depends on the coefficient of wavefront aberration, the coefficient of Z12 is not zero, that is, there are 4th order (the order of / 0 is 4th order, n ~ Fengqian) Those who do not express high-order astigmatic aberrations. As shown in the simulation results of…, when ^ is zero, that is, when 65 200400540 VOm is again, it is the magnitude of the coefficient B of the ninth term of the fourth-order 0θ component Irrelevant, the wave surface in the pupil plane becomes the same chaotic pattern in either direction. From this we can see that 'Zΐ2 = 0, that is, in the upper part of Fig. 6 (Fig. M ~ Fig.) I3 right contour pattern composed of a plurality of concentric circles. I3 right Z 丨 2 〇 'Here, because the line width difference between the vertical and horizontal lines as the purpose: control (adjustment) is difficult, wavefront aberration is used for any evaluation point. : The k coefficient of the S 12 term obtained from the f measurement result is assumed to be zero. As for the actual projection optical system, the system's task in the field of view-an evaluation Point, the Z12 coefficient of the twelfth term of the checker polynomial of the wavefront aberration is usually not zero: k kinds of assumptions can be said to be true. Also, in this embodiment, the so-called coefficient of the ninth term is set as follows That is, at each evaluation point, the amount of change in the coefficient Z9 of the calculated operation term is calculated based on the line width difference ΔL of the vertical and horizontal lines and the coefficient Zi2 value (size) of the twelve terms of the Chanelian brother. The target value ^ ......... Regarding the basis for this calculation, we will describe p. 2 33) () ^ The target of the change amount of other terms = the wavefront aberration expressed by the following formula ⑽ The target value of the change amount 4 Q, = ·: [Pn \ 66 200400540 In the above formula (14), each element is a column matrix (column vector) of the following formula (15J, (152), .... P] ,, P, 12, ......, V (n ..., Π5η) 3? Row 33) is a column

Ρ2- • V 〇 ^ 1,8 0 z SS heart 1,9 ^ 1,10 0 ^ 1,37 0% · _0 · ^ 2,8 0 ^ 2,9 s = r2 ^ 2,10 0 ^ 2, 37 0 (152) (15,)

^ 33,1 '0 ^ 33,8 0 ^ 33,9 ss ^ 33 of 33,10 0 ^ 33,37 _ 0. · (15λ)

From the above formulas (15ι), (152) ..... (15J also knows that

Pl ', P2' ........ ρ: can be regarded as their evaluation points (measurement points / gram polynomial coefficients other than the ninth term are all zero, can be regarded as 々 $ number is rj (j = i ~ 33) of 37 rows] column matrix matrix ⑼ vector ^ brother 9 J | Therefore, 'If the matrix r is used to represent the key 67 of the parameters in each adjustment at this time, use the aforementioned matrix ο, Q, = The display system of 0 · P,) is established ... · · · · · (1 / λ \ in the above formula () 6) t, ρ, system — (from the matrix of the table ADJ1 ADJ2 ρ1 • (17) ADJm is below- In step 122, the main control device solves the above formula (16), finds the moment system and uses the Xiaoping method to sacrifice early P (the operation of the formula (18) by each adjustment amount. 冓 成). That is, perform P, = (〇T · 0) 1 · 〇τ ^ ... (18. In the next step 124, the P of the person #, that is, the main control device 50 of the mountain towel is calculated according to the above. After adjusting the adjustment amount (ADJ1 ~ Fine 5, ADJ19), as described above, control the movable lens 13 A. DT ^ 1 5 to adjust each part specifically, adjust the projection light stem 1 ',, and, and end a series of processing. Also, about Wafer position = type adjustment amount Α_ ~ adjust i Amu, use Later description: The position control of wafer stage WST of the magical exposure%, the story is first memorized in the dove or _ set 44. By this, within the field of view of the projection optical system PL = wavefront aberration of 33 evaluation points, specifically, The ninth term coefficient of the Chase polynomial is adjusted only by the projection optical system pL with 0 changes. As a result, using the adjusted projection optical system PL ·, the vertical line (V line) pattern and the ray line (H line) are adjusted. The circuit diagram 68 on the reticle r with the patterns mixed is transferred to the wafer w. Therefore, in order to make the vertical line (v line) pattern and horizontal line (H line) pattern image line width difference (vertical and horizontal) The line width difference of the lines is approaching the design value, for example, 'corrected so as to approach zero. Here is a detailed explanation of the reason why the above-mentioned projection optical system PL can be adjusted to correct the line width difference of the vertical and horizontal lines. The lower part of the figure (Figure 6D to Figure 6F) indicates that when the coefficient of the 12th term of the Chagneck polynomial of the high-order astigmatic aberration term is ζ12 = + 20ηιλ, according to the Chagnek of the low-order spherical aberration term Wavefront change in the pupil plane of the ninth term coefficient & of the polynomial Where (simulation results). Wherein, based on FIG Z9 = -20rn λ represents the case, based on FIG. 6E showing z9 = 〇m and the case, based on FIG. 6F showing Z9 == + 2〇m and the case.

1 If the 2nd component is not zero, if you make the 9th item

69 200400540 Here, 'the vertical line (v-line) pattern on the reticle has spatial frequency components in the horizontal direction', so from the vertical line (v-line) pattern, diffraction light is generated in the left and right directions' the horizontal line (H-line) pattern is in the vertical direction The direction has a spatial frequency component, so diffracted light is generated from the horizontal line (H line) pattern 'in the up-down direction. Therefore, as described above, when the signs of the 9th item and the 12th item are equal (in the case of FIG. 6F), in the left and right directions where the phase change is large, the contrast of the vertical line pattern image that produces the diffracted light becomes lower, and the line width becomes thinner. . In contrast, in the phase change, the contrast of the horizontal line pattern image of the diffracted light is almost unchanged.

Low, so the line width is about the same as the design value. As a result, the line width difference between the vertical and horizontal lines becomes a negative value.

In contrast to the above, when the sign of item 12 is positive, if the number 4 of item 9 is negative, the phase change of the 12th item from the left and right of the pupil is positive, and the phase change of the 9th item is negative, so Weakening each other, the phase change of the ^ member and the phase change of the ninth term increase negatively above and above the pupil. This sigmoid, for example, becomes the wavefront distribution in the pupil plane shown in FIG. 6D. $ :: shape, the horizontal coil of diffracted light is generated in the up and down direction of the phase change. The right ^ contrast becomes lower and the line width becomes thinner. In contrast, in the j direction with a small phase change, the contrast of the vertical line pattern image that produces the diffracted light is almost inferior to the line width, which is roughly the positive value of Koko + Bo η and α. The values are the same. Eight ,, σ, the difference between the width of the vertical and horizontal lines becomes the pupil ^ As mentioned above, when the 9th and 12th terms are not zero, the wavefront disorder of the light direction and the left and right direction is based on the 9th item The sign differs from the sign of the value of 1. The focus is on this point, and the 9th item (low-order spherical aberration component) which is easier to adjust under the ^ item "70" 200400540 'This can adjust the vertical and horizontal lines The line width difference. The difference in direction between the phase changes of the ninth term ㈧ above is the 4th order θθ component and the 12th term (the order of 々 is the 4th order 2Θ component (cos20 component)) is reduced by P If the order of the 4th term is taken into account, the defocus term of the component is 2: coefficient Z4) and the 5th term is 0. The low-order astigmatic aberration of the 2nd order ^ component (⑽2Θ component) Worker: Silk a) Relative comparison work, it is easy to understand. The sign of the main item * is positive. For example, as shown in Figure π, the phase in the left and right direction of the pupil is positive, and the phase in the up and down direction is negative. On the other hand, * the sign of the 4th term is positive, for example, 'as shown in the% chart, the phase of the pupil peripheral edge is positive' when the sign of the 4th term is negative For example, as shown in Figure 7A, the phase of the pupil's outer peripheral edge is negative. Therefore, when the signs of the 5th and 4th terms are equal to the size of the inch and the figure 7F, the light in the left and right directions The phase change is large, and the phase of the up and down direction Π is small. On the contrary, when the symbols of the 5th and 4th items are pretty; ^ 荣 # ^ 仃 观 系 is not the same as the 7D picture-like situation, in the eyes Left and right: The phase change is small, and the phase change in the up and down direction is large. Θ When the fifth member of the field is not zero, the vertical focus and vertical line patterns are based on the coefficient z5. The value varies, so if the fourth item is changed, the difference in line width (line-finding difference) of the vertical line (V line) pattern toilet green m and the line of evil (line line) pattern image will be based on the defocus ( That is, the change of item 4 and item 4) is generated. That is, when the line width of apricots changes due to defocus, it must be as shown in the CD-focus line diagram of Figure 8 as shown in Figure 8. Focus position other than zero, 1 corresponds to # 5 of the best focus position of the vertical line pattern and horizontal line pattern of the 5th item, square A, A, produces a vertical line pattern image (V) and a horizontal line graph The difference in line width of image (H). This is the astigmatism aberration of 2 Θ components that are generally seen. 71 200400540 The component change (defocus) at time 〇0. For vertical line pattern image and horizontal line pattern image · Line width It can be seen that when the fifth term of the astigmatism aberration component of the lower order is not zero, adjusting the fourth term indicating defocus can reduce the line width difference between the vertical line pattern image and the horizontal line pattern image. Figure 9 shows the use of a KΓρ laser with a wavelength of 248 3nm as the light source. Illumination is 2/3 of the ring illumination condition with σ = 0.75, and the numerical aperture of the projection optical system PL (in the case of NAHM, transfer the aforementioned measurement) An example of the experimental result of the line width difference of the vertical and horizontal lines of the photoresist image line width measurement result obtained by using the pattern on the reticle RT. Figure 10 shows the part of Figure 9 in more detail: Z12 = 40 mA, 2 () mA, and 0 mA (part of the previous 3 paragraphs), and Figure η shows the details in Figure 9 The part of Zi2 = -20mA, -40m (the part of the next 2 paragraphs). In the graphs and η graphs, the & axis system of each contour graph indicates the coefficient of the fourth term. 24 'The vertical axis system indicates the coefficient of the fifth term. From these descriptions, it can be seen that the 9th graph indicates Inside the car's perimeter and step 20 "U steps make the combination of the 9th coefficient & the 12th coefficient " 12-year value of each z9, Zi2 ai, and & and & respectively change 1㈣ The relationship diagram. · The shirt line attached to each area in the 9 figure is specifically shown in Figure 9 less α-the line width of the vertical and horizontal lines below the figure = the value ^ indicates the vertical line ( V line) The line of the pattern image Η The line width is thick. The area where the line width difference is negative indicates that the line width of the vertical line (V line) pattern image is thinner than the line width of the horizontal line (Η line) pattern image. Figure 2A shows the 9 --- Gaiyou contour map shown in the upper left of Figure 9; Figure ⑶ shows the CD-focus map corresponding to CC line 12A 72 200400540. Also, Figure 12B shows Take out the contour map of Zi2 = 4〇m; l, Z9 = 〇m, which is not in the middle of the slave on the top of Figure 10. Figure 12D shows the CG-focus line map corresponding to 12B and line. From these figures, we can know The contour maps in Figure 9 show how to change the v-line according to the changes in the defocus term (coefficient Z4) and the low-order astigmatic aberration term (coefficient Z5) under some conditions of Zi2 and A. A diagram showing the difference in line width between the pattern image and the H-line pattern image (hereinafter, both are referred to as "VH difference"). When the value of z12 is zero (from above in Figure 9 and $ 3 side by side with paragraph 3)

Contour map), the value of z9 varies from map to map, but as shown in the maps in the lower section of Figure 10, no matter which focus position is in which map, if the value of B is zero, no VH difference will occur.

In contrast, when the Zu value is not zero, for example, # Zi2 = + 4〇m, and from the upper graphs in Figure 10, it can be seen that if s z5 = 〇, V will be caused by Z, 2. The optimal focus difference between the line pattern and the H line pattern. Although the VH difference of the line width varies due to the value of Z4, at this time, the value of the VH difference of each Z4 varies due to the size of Z9. In order to understand the influence of the vh difference & clearly, a system equivalent to the high-order astigmatic aberration term—25 is provided. The low-order astigmatic aberration term (coefficient 25) is used to correct the high-order astigmatic aberration term. The optimal focus difference between the V-line pattern and the stern line pattern of item 12 is the case where there is no focus effect of the line width difference. The VH difference of the line width is affected by the heart value. When the value of A] is positive, the line width of the ν line pattern image is thinner than that of the Η line pattern. Conversely, when the value is negative, the V line pattern image is The line width is the thick line width of the car crossing line pattern, and it can be seen that it can prove the content explained previously using FIG. 6. 73 200400540 The results of experiments conducted by the inventors used better focus difference (under normal lighting conditions of illumination ° 0.4, line width 0.72 / zm line and gap pattern (L / S pattern) and line width l4 # mL / s pattern, the difference between the better focus positions), change the size of the spherical aberration term Z9 from —〇18 # m to —〇〇2 # ηι 'to confirm that the VH difference of the line width can be changed from 27nm is reduced to 7 ~ 8nm. From the above explanation, it can be proved that, as mentioned above, the value (size of each evaluation point in the field of view of the projection optical system according to the line width difference (VH difference) of the vertical and horizontal lines and the 12th term coefficient of the Zernike member formula (z12) ), Perform a predetermined operation, and π output the target value of the ninth term coefficient & the amount of change qq ........ Γη (η == 33), and according to the calculated target of the ninth term coefficient of change Z9 Value, to adjust the projection optical system, so that the line width difference between the vertical and horizontal lines can be adjusted. However, in the exposure device 100 of this embodiment, when manufacturing a semiconductor π component, the reticle R for manufacturing a component is carried on the reticle table RST, and then the reticle is adjusted and the so-called baseline measurement, Preparations such as wafer measurement such as EGA (Enhanced Global alignment) Regarding the above-mentioned preparations of the reticle alignment and baseline measurement, for example, 'Details are disclosed in Japanese Patent Application Laid-Open No. 7_176 and corresponding publication, US Patent No. 5,646,413, etc., and secondly, regarding ega, \ It is disclosed in Japanese Unexamined Patent Publication No. 6 1-44429 and the corresponding U.S. Patent No. 4,78M17 #. The disclosures in each of the publications and the aforementioned U.S. patents corresponding to this publication are referred to as described in this specification. One part 74 200400540 and then 'based on wafer alignment results. In addition, because the exposure during the exposure process is the same as that of the normal scanning exposure device, so it is the flowchart of the exposure device diagram in the above-mentioned step scanning method: the second embodiment. The adjustment method shown, "Then use hunting by Besshudi 4

The projection optical system PL that is adjusted by the method of v ~ is adjusted according to the calculated position and method of ^ K W of A Γ 1A, and this control is performed by M Ϊ 6 ~ 8. By virtue of the fact that the electrical pattern image and horizontal line pattern image formed by the reticle R are from the palace to ^,,,,,,,, and R lines are reduced, these images (latent images) are formed Every irradiation field on the sun.

From the above description, it can be seen that the adjusting part of this embodiment is composed of movable lenses 13! To 135, wafer A WST, and light source Η, and the position of the movable lens 2 Θ 方向 (or its illumination light I and W changes I and W). The amount of wavelength shift becomes the adjustment amount, and the above-mentioned adjustment section, the correction controller 48 that drives the imageable performance, and the driving stage WST 2 = the moving section (not shown) constitute the image formation state adjustment device. The dream is constituted by the master device 7 device 50 to control the image formation state adjustment device. However, the structure of the image formation state adjustment device is: Ming Ming, :: For example, as far as the adjustment part is concerned, Contains only movable: 5 k because the imaging performance (variable aberrations) of the projection optical system can be adjusted even in this case. In addition, the present embodiment uses the main control device 50 (calculates on the imaging wafer) The alignment system and system ALG of the photoresist image of the measurement pattern formed and the photographic signal according to 75 200400540 ALG system ALG, the vertical pattern, pattern, and horizontal pattern light included in the measurement pattern Line width of blocking image) to form a line width measuring device As for the line width measurement device, for example, a dedicated measurement device (SEM, etc.) provided outside the exposure device 100 may be used. Also, in the above description, IV, The measurement of the wavefront aberration is performed using a wavefront aberration measuring device Kawasaki based on a space image formed through a pinhole and a projection optical system PL, but is not limited thereto. For example, US Patent No. 5,978, 〇85 The measuring hood with a special structure revealed by the catties, etc., is sequentially printed through a set of repairs, 'ten holes and projection ~ optical subsystem', each pattern on the hood is printed for each measurement On the substrate, and through the projection optical system without transmitting the condenser lens and pinhole, the reference pattern on the hood can be printed on the soil plate, and multiple measurement patterns can be obtained as a result of each printing. The photoresist image of each reference pattern of the photoresist image is used to measure the position shift and calculate the wavefront aberration by a predetermined calculation. As described above in detail, if the method of adjusting the projection optical system PL according to this embodiment is used, ,then Adjust the projection light H system PL to control the high-order astigmatic aberration (item 12) of the _ projection optical system PL, which is not easy to adjust. The width of the aforementioned vertical and horizontal lines is generated to control the easy adjustment. Low-order spherical aberration (item 9). Therefore, it is possible to freely and surely control the line width difference between the vertical line pattern image and the horizontal line pattern image, which is difficult to learn. ▲ And if the exposure equipment according to this embodiment is used, ι〇〇 and the exposure method of the tongue by wavefront aberration! Measuring device 80 energy measurement projection optical system 'system PL H surface ϋ' '1 measurement reticle & measurement pattern is through projection 76 200400540 System PL, transferred to the crystal ... After development, the photoresist image of the pattern for measuring the wafer formation is developed by the main control device 50 to close the case: ^ Let's take a picture. 'Calculate the line width of the photoresist image of the vertical line pattern and horizontal line pattern with a sufficient amount, and the horizontal line pattern based on the shooting signal.

The main technical device 50 develops a high-order image of the surface aberration check polynomial of ... measured by wavefront aberration measurement: when the aberration term U 1 is not equal to zero, it is based on the The value of item 12 (factor Zl2) and the difference (line width difference) between the i-th line width of the vertical line pattern image and the second line width of the horizontal line pattern image measured above (line width difference), before use: The image formation state adjusting device 'controls the magnitude of the low-order spherical aberration term (second optical characteristic) of the 帛 9 term of the Zernike polynomial which affects the line width difference due to the interaction with the 帛 12 term. . That is, the image formation state adjustment device is used to control the magnitude of the low-order spherical aberration that is easy to adjust, thereby controlling the aforementioned line width difference caused by the existence of the high-order astigmatic aberration that is not easy to adjust. Furthermore, the circuit pattern of the reticle R is illuminated with the illumination light EL, and the circuit pattern is transferred onto the wafer W through the adjusted projection optical system PL. This makes it possible to effectively reduce the line width difference between the transferred images of the vertical line pattern and the horizontal line pattern. Moreover, in the above-mentioned embodiment, the case where the 12th term of the Zernike polynomial (high-order astigmatic aberration term) of the wavefront aberration is developed for the first optical characteristic system, and the Zernike polynomial term of the second optical characteristic system is expanded. The case of 9 terms (low-order spherical aberration terms) will be described, but the present invention is not limited to this. For example, in terms of the first optical characteristic, it is also possible to measure the order of rabbit p which is the same as the item 12 above (the order of p is a cos2 0 component of the order 4 77 200400540). The white number is a 2Θ component of the fourth order. (Sin2g ingredient) Item 13. In this case, as far as the first learning characteristics of Fandidi 2 are concerned, it is still possible to use the same 9th Jg and 3 items as the above embodiment. Due to the interaction between item 9 and item 13, on the reticle, consider the vertical line (V line) and horizontal line (H line) described by a pair, so use 45. & The line width difference between the oblique line pattern image i and the oblique line pattern image of the sire 2 of this line's father is affected. Therefore, as in the above-mentioned embodiment, when the 13 terms of the Chaney polynomial and only the brother of A are not zero,

According to the value of the U term (coefficient Zl3), the difference between the line width and the second line width of the measured line width of the ^ line graph, the second line width of the slash pattern image, and the second line width (line width Difference), using the aforementioned image formation state adjusting device to control the magnitude of the low-order spherical aberration term of the 9th term of the Chase polynomial, thereby controlling the aforementioned line width difference.

Others can also consider the optical characteristics of 帛 1 as the second and thirteenth rotational symmetry components of order (m-4) other than the above-mentioned 12th and 13th terms of the Chakker polynomial of expanded wavefront aberration. 2 The optical characteristic is regarded as a rotationally symmetric component of the same order as the second-order rotationally symmetric component described in item 7 (2). Foot sample-come 'When the size of the first optical characteristic is not zero, it is based on the difference between the line widths of the two patterns of line patterns orthogonal to the measurement based on the size of the first light and the two orthogonal directions of the measurement.帛 2 Optical characteristics, the projection optical system PL is adjusted, and it is estimated that the same effect as that of the above embodiment can be obtained. And 'as far as the combination of the first optical photonic characteristic and the second optical characteristic that affects the aforementioned vertical and horizontal line difference width (νίί difference), in addition to the 9th and 12th terms of the Zernike polynomial described above In addition to the combination, there are also the 6th (coefficient Z6) and 18th (coefficient ZJ, 13th (coefficient Zi3) and 78 200400540 18th (coefficient Z18), 12th (coefficient zi2) and 17th ( Coefficient Z17), etc. The inventors etc. 'can effectively determine whether the combination of the Zernike polynomial term (Chag's gram term) that develops wavefront aberrations is the cause of the VH difference, so they are used to specifically find the cause of the VH difference The combined simulation is shown in Fig. 13. Figure 13 shows the graph when the light source is an ArF excimer laser (wavelength is 193nm), the numerical aperture of the projection optical system pL (na) = 〇68, the illumination σ -0.85, and the 2/3 ring zone. Under the lighting conditions, the isolated line (2 # interval moment) with a line width of 140nm (including the hood deviation ·· + 4 Xinjiang) is used as the line pattern with the conversion value on the wafer, and the transmission rate is 6%. Half-tone phase shift hood (reticle), processing the line width of isolated lines on the wafer The calculation result of the parent term between the aberrations at 100 nm. In this figure, Zi (i = 4 ~ 20) represents the first term of the Zernike term. In this figure, the oblique The upper right side of the boundary line is the size of the parent line of the aberrations (Chagneck term) and the horizontal line, and the lower left side is the size of the line of the father cross term of each aberration (Zanek term). From this section The table in Figure 13 shows that the intersection of the combination of item 9 (Z9) and item 12 (Z12) is 759 on the horizontal line and _-759 on the vertical line. The signs are exactly the opposite. ', His item 6 ( Z6) and item 18 (Z18), item 13 (Z13) and item 18 (Z18) 'item 12 (Z12) and item 17?), Etc., the intersection of vertical and horizontal lines and the opposite sign are also opposite , Become the original head of the VH difference. As for the combination of the first optical characteristic and the second optical characteristic, although the above-mentioned combinations of the various search / conversion terms are considered, it is not limited to this, and the first optical characteristic may be regarded as an astigmatic aberration Consider the second optical characteristic as spherical aberration 79 200400540. In this way, if the interaction between the astigmatic aberration and spherical aberration is considered to affect the line width difference between the vertical line pattern image and the horizontal line pattern image, astigmatism exists when the measurement results of the optical characteristics of the projection optical system are present. In the case of aberration, the line width difference between the astigmatic aberration and the measured line patterns in the orthogonal 2 axis directions is different. In order to control the spherical aberration, the projection optical system is adjusted to control the line width. difference. It is clear if considering this situation, but the optical characteristic measuring device of the projection optical system PL is not limited to the wavefront aberration measuring device, but may be used to measure the spherical aberration and image of the projection optical system pL. Astigmatic aberration and other so-called 5 aberration devices of Saidakura. For example, there is a so-called aerial image measuring device, which is formed on the wafer stage WST to form a slit-shaped or rectangular-shaped opening pattern, and it is a predetermined pattern formed by the projection optical system π. The space image of the measurement pattern is measured to scan the opening pattern. The photoelectric element is used to detect light passing through the opening pattern. Further, in the above embodiment, the wavefront aberration of the projection optical system PL · is directly measured using the wavefront aberration measuring device 80, but it is not limited to this because the twelfth term (the coefficient zy is a high-order astigmatic aberration) Therefore, for example, it is also possible to obtain the preferred focus positions of a plurality of L / s patterns (or η-line patterns with different directions) with different periodic directions. The astigmatic aberrations obtained from this result are the most suitable. The Xiaoping method et al. Calculated the unterminated coefficients of the linear combination of the low-order astigmatic aberration term (coefficient B) and the high-order slit aberration term (coefficient ZO 2), which were used to determine the 12th term ( The approximate value of the coefficient Zιζ). The above-mentioned different focal positions of the L / sw case with different periodic directions are obtained. The position of the wafer is changed by changing the optical axis direction of the wafer. 200400540 〃 is formed in each irradiation area on the circle obtained by printing the above pattern on the wafer), and the above-mentioned aerial image measuring device can also be used, and the light of this aerial image measuring device can be changed. The position of the axis direction, the space plane of the above figure Γ is measured: Calculated based on the measurement results of this aerial image. Muji: The same structure as the above-mentioned embodiment is used to configure the line width measurement. The measurement is not limited to a photoresist image, and it can also be formed on a wafer. The line width of the latent image or engraved image of the aforementioned vertical line pattern and horizontal line pattern. Alternatively, the line width measurement device may be constituted by an aerial image measurement device. In this case, for example, on the image plane An m-image image measuring device that forms a vertical line pattern and a horizontal line pattern is used to measure the line width of the space image. That is, the simultaneous line image :: measurement of line and line width. ^ $ In the above embodiment, In order to simplify the explanation, the difference between the line width of the image of the vertical line pattern and the horizontal line pattern is substantially zero: it will be explained when the width difference is controlled. It is not limited to this. If it is a pattern in the orthogonal 2 axis direction:,, then it has nothing to do with the line width (even if the line width is: 2: the control of the line width difference between some images. For the control of the line width difference The language approached the line width difference of the pattern image of the line width difference on the design correctly. The shadow optical system is installed in the state of clothing) to adjust the projection optical system, but for example, in the manufacturing process of a device (especially a projection optical system), it is also possible to use a single unit to adjust the projection optical system before entering the exposure device. ° In exposure, however, the two main lines of the orthogonal pattern of the direction of the car are caused by the aberrations of the projection optical system. Also consider: 81 200400540 The pattern drawing error caused by the film. & Amp In order to reduce the difference between the line widths of the orthogonal two-axis line pattern images caused by this main cause, the adjustment method, exposure method, or exposure device of the optical system of the present invention can also be used. The extended case is related to Similarly to the above embodiment, according to the measured optical characteristics of the first line, for example, the twelfth worker value of the Zernike polynomial, the line width of the known first line pattern, and the line width of the second line pattern orthogonal to this. Difference (Daylight Error): Due to the interaction with the first optical characteristic, the image line width of the j-th line pattern on the image plane and the second line by the projection optical system and Yang Cheng

The second optical characteristic of the difference in line width (line width difference) of the pattern image is affected. For example, in order to restrain the size of the 9th term of the Zernike polynomial, the projection optical system is adjusted. Therefore, the projection optical system is formed on the image surface. The difference (line width difference) between the line width of the first line pattern image and the line width of the second line pattern image described above can be controlled freely when the line pattern is orthogonal due to the day-to-day error of the pattern on the reticle. Line width difference between each other. It can also be seen from the description on 1 that the poor information of the aberration of the method of the projection optical system can be obtained, and the information of the case image can be further obtained.

Jin-: When adjusting the projection optical system based on this information, use: Niek polynomials, which can be used to change the characteristics of other projection images in a plurality of Zhani members who expand the wavefront aberration series. The characteristics of the interactive projection and projection projection characteristics are combined with any combination of grammar and projection items), and the sensitivity is used to adjust the aforementioned projection optical characteristics and grammar M w to change the aforementioned projection image characteristics (not considered in the conventional knowledge) The cross-terms of the combination of the characteristics of each other's fairy shadow, s 古 馆 +, portrait ^, and a combination of Niek members), consider the Zanike senses to adjust the projection optics, system, and practice Know the aberration components that are difficult to adjust, such as 82 200400540 'Can also adjust high-order aberration components, etc., and can adjust the projection optical system with a better formation state of the pattern image. In this case τ, when the pattern includes a line pattern, as far as the characteristics of the aforementioned projection image, the characteristics including at least the line width of the line pattern can also be considered and the sensitivity can be checked. χ, in this case, the adjusted ix W optical system and system can also be used to transfer circuit patterns onto wafers and other objects. This also enables high-precision pattern transfer. -4 疋 Not only the line width difference of the vertical secret line, but also the isolated line pattern like

The line width is also affected by the amount of defocus. Here, # 明 者 等 is performing experiments in order to obtain the aforementioned CD-focus curve. (slice level), to find the line width deviation CD at the position of + 0.15 // m defocus. In terms of exposure conditions, it is assumed that the light source is an ArF excimer laser (wavelength: 193nm) 'Numerical aperture (na) of the projection optical system pL = 178, illumination σ = 0.85' 2/3 ring-shaped illumination conditions, object pattern formation Calculate the exposure amount at the half-tone hood (reticle) 'using wafer conversion values, an isolated line with a line width of 100 nm (2 μm pitch), and with no aberrations and no defocusing, for example , Input the 50m aberration into the optical term to the 37th term to find the Zernike sensitivity

The computer used in the simulation of each Zanek term is shown in Figure 14 if it is from the first. In Fig. 14, z.1 (i = 37) on the horizontal axis represents each Zanek term. In Bai Zhizhi ’s Zernike Sensitivity method (herein referred to as the “ZS method”), Zhongtai Datian 4; ^ This is expressed by the following formula (19)

Sensitivity (hereinafter,) Si (i = l to 37), and the measurement point (hereinafter referred to as the measurement point, etc.) called "Zanek Sensitivity" or "zs" for short. The size of the gram term cn, i (= the linear combination of the coefficient zd represents the line width deviation ΔCD. Also, hereinafter, Cn, i is simply referred to as the Zanek term component (choke gram component) of each measurement point. 37 ACD = ^ SiCni ··· (19) / = 1 However, it can be seen between the calculation result of the ZS method using the above formula (19) and the method of directly calculating the spatial image by providing the wavefront aberration of Shida. The deviation shown by the curve in Fig. I 5 is that the calculation error of the ZS method of m formula (19) is too large. Therefore, the inventors considered to make the CD-focus in the quasi-plane using the focus and line width as the coordinate axis 2 The method of estimating the line width by moving the curve. As mentioned above, because the line width prediction error of the method of △ CD is directly expressed by the linear combination of the components of the 圼 圼 term, the calculation of the 查 CD sensitivity and the calculation of △ CD are large. Add one step of parallel movement of CD · focus curve between

: According to the CD_focus curve after this step (parallel movement) is performed, the method of ΔCD is adopted. For example, the movement of the CD-focus curve refers to the movement of the focus direction 各 and the movement of the line width (⑶) in the direction of the measurement point n (D.). Y ^ f (x-an) + / 5n ^ CD .; Tl focal curve, for the newly created function, calculate △ cD. (Prediction method) The following is an example of the prediction method of the present invention, the CD-focus curve preview 84 200400540. The implementation of the measurement method is shown in the flow chart (Figure 16 and 24 ffi) along with the process flow, and with reference to other The drawing is illustrated. In the present embodiment, regarding each measurement of the wavefront aberrations of the exposure device 100, the point n (n = 1 to 33) 'is a linearity using a plural number including a check component cini. Combined with the Zanek sensitivity method, to predict the characteristics of the pattern projection image-the CD-focus curve and line. As for the components Cn > i of each grammatical term, it can also be obtained in advance by the imaging simulation on the simulation battery 46.

As described above, the value obtained by the measurement by the wavefront aberration measuring device 80 may be used. 'First, in step 202 of FIG. 16, the input device through the simulation computer 46 will include the optical conditions at the time of actual exposure (for example, the illumination light EL < wavelength', that is, the exposure wavelength (and Types of light source 16, etc.), the highest λNA (numerical aperture) of the projection optical system PL, ❹NA (in this embodiment, the numerical aperture set by opening 15 sigma aperture during exposure), coherence coefficient σ value (illumination σ) or illumination NA (the numerical aperture of the illumination optical system), and the illumination conditions of the reticle (the pupil surface of the illumination optical system

The exposure conditions of the light amount distribution of the illumination light EL, that is, the shape and size of the secondary light source, etc.) are set in the simulation computer 46. At this time, the imaging simulator "on the simulation computer 46" has been started, and on this screen i, the exposure condition setting screen of the imaging simulator is displayed. The operator and others set the optical conditions used in the actual exposure in accordance with the setting screen. #When setting the lighting conditions, 'relevant patterns are transferred to each measurement point n (n = 1 ~ 3 to take the pattern on the graticule). Information about patterns such as shape, size, etc. is also set. Regarding the pattern of the pattern, 5 for example, there are isolated line patterns, line and gap (L / s) patterns, lines 85 200400540, which: the type of pattern (including whether it is The phase-shifting pattern and the selection of major patterns such as the line width, length, and spacing of the pattern are determined based on the evaluation items that can be evaluated. For example, = then Yi 2 is the same as * 'If the vertical and horizontal line width difference is used as the evaluation item "If you have a reticle formed by the orthogonal line patterns of the reticle ^ shown in Figure 5 above, you must set information about the pattern ^ ^ _ Yano.

Here, in order to explain ‘说明 each measurement point’, the same-sized isolated relay is used: the standard mark is set, and the pattern information on the reticle is set. In the next step 204, the imaging simulation is performed to create a CD · focus curve when the projection optical system has no aberrations. Specifically, the operator or the like instructs the simulation computer 46 to create the CD_focus curve when the projection optical system 2 PL has no aberration through the input device. In response to this indication, the second proposed computer 46 is under the exposure conditions set in step 202 above, 'assuming that the projection optical system has no aberrations, that is, in (that is, the factory of formula (3), The Niek term component Zi (i = l ~ 37, that is, Cn "is all set to 0,

The line width versus defocus amount of the line pattern is made by the imaging simulator: that is, a CD-focus curve is made. In the next step 206, the simulation computer 46 adjusts the created CD-focus curve using a 10-degree function shown in the following formula (20). y = Cax10 + Cbx8 + Ccx6 + Cdx4 + Cex2 + Cf (2〇) In the formula, X is the defocus amount, and y is the object pattern corresponding to the defocused shape ..., in step 202 above, set the pattern information Line pattern) image line width, c ~ Cf are coefficients of each order of the 10th order function. From this formula (20), it can be seen that the Xi function is a function composed of even-order terms of the 2nd to 10th order. Fig. 86

ZUU4UU ^ 4 (J is the result obtained if the table should not be adjusted <

The figure shows an example of the number of functionists and an example of the adjustment error of line 17B. The adjustment error of the CD-focusing curve obtained by the second image drawing is less than ± 0.02, and the accuracy is 5 weeks. / In step 2G8, the operator and the like use the simulation computer 46 to produce S々. μ · Take a Niek term to the Zanek sense of political focus: Behind the 'For example, if you only input 1 kind of aberrations expressed by Zanek polynomials to make a CD-focus curve, you can know that Phases-distribution. Here, in this allocation, the change in the preferred focus position is shown. The M-th graph shows the search sensitivity Sai. In Fig. 18, the horizontal axis to 37) indicate each Zanek term. As shown in Fig. 18, the sensitivity to the defocus amount becomes the Zaneck term z4, z9, Zi6, z25, Z36, z ", or rotation injury or z5, zi2, Zn, Z21, z28, z32, etc. The second-order rotation becomes wound, that is, only the even number of 0 components. The other (odd number of 0 components) is that the Zanek sensitivity is zero. Therefore, the CD-focus curve, with regard to the shift of the focus direction, the odd number of 0 is not affected. The cross terms are not affected. _ Fig. 9 shows that from 50m to 50m λ, the 10th λ spacing is used to make the 9th (Z · 9), 12th (2.12), and 16th (Z · 16) The amount of movement in the focus direction (aj's change) of each of the three types of grammar items when moving 11 points. In Figure 丨 9, it is obtained by using the least square method according to the amount of movement of each 11 points The slopes of the straight lines are also shown together. That is, the slopes of the straight lines become the values of the sensitivities of the Zanike terms. In Fig. 19, only the 9th item (Z.9) and the 12th item are representatively shown. Item (Z · 12), Item 16 (Z.16) of 3 87 200400540, but for other Zanike terms, if the amount of aberration is also within this range, it can be confirmed that the line is almost completely maintained As can be seen from Figure 19, because the correlation coefficients R2 and 1, the usual Zanek sensitivity is calculated. In order to save time, you can also enter a 1-value aberration to obtain the slope of the straight line. Here, to confirm β tolerance and derivation of more correct values for the purpose, from a 50m λ to 50m, and with a distance of 10m #! i point focus direction movement (qj, using the least square method to calculate the slope of the straight line ( That is, the value of Zanek's sensitivity).

Its-man, in step 2 10, simulate the use of the Chagnek sense of each Chagnek term by a computer for private use | Sai, and the Chagnek term of each measurement point η㈣ ~ 3 ”wounds ~ 3 3) , Using the same formula (21) as the method of obtaining the comprehensive focus difference (tfd), astigmatic aberration, and spherical aberration, the following formula (21) is used to find the n-point (n = 1 3 3 ) "The offset of the focus direction of the CD-t focal curve of J on the ear is also« η ° 37 an ^ Y, SaiCni (21) / = 1 4 疋 In each Zanike term, from ~ 50m 50m again, use 10m

The door distance is input by 11 points of aberration, and the relationship between η (the line width of the preferred focus position) and the aberration is obtained from its distribution. Whether the amount of aberration is positive or negative, the effect is the same: 'if the aberration increases Then, from the observation of the proportional relationship above the image: Deterioration '/ 5 η can be expressed by each combination: a linear combination of ¾ grams of component squared. Here, in the next step 212, the operator uses a simulation computer under the predetermined exposure conditions. #Calculate each check by aerial image.] F: Zanek sensitivity s of the square pair line pattern line width ^. In the calculation of the line width variation of 11 points obtained by using a 6 ten different, it is assumed that 88 200400540 P white function 'approximated by the method of least squares, as shown in Figure μ, this observation To the function mounted on y == sx2. In Fig. 20, although only three types of the 6th item (26), the 7th item (27), and the 9th item (ζ · 9) are representatively shown, with regard to the other Zanike terms, Confirm that you can also use the quadratic function to express. Also, Fig. 21 shows the sensitivity of each Zanek term (S / 5 1 37). In Fig. 21, Z i (i = 1 to 37) on the horizontal axis represents each Chak / Rick term. As shown in FIG. 21, for the line width change, the odd-numbered 0 component and the even-numbered Θ component have sensitivity. Next, in step 2-14, the simulation computer 46 reads the Zernike term components Cn, i of the measuring points from the memory, and uses the following formula (22) to find the CD- Deviation of the focus curve 37 n (22) / = 1 Next, in step 216, the simulation computer 46 uses the αη obtained in the above step 210 and the 10,000 obtained in the above step 214. Based on the following formula (23), the CD · focus curve of the measurement point η (η = 1 to 33) is obtained. The thus obtained CD-focus curve is a CD-focus curve predicted at each measurement point η when the aberration of the projection optical system PL of the measurement point n is considered. However, the deformation of the CD-focus curve obtained here has not been considered. yn = ca (X-α. 1. + Cb (x— α n) 8+ c “x — α n) 6+ Cd (x — α j4 + ce (x— an) 2+ Cf + cold n… (23 ) Again, when /? N is found, the following formula (24) can also be used instead of the above formula (22). 37 37 / = 1 y = l 89 (24) 200400540 The above = formula (2 4) is to consider each other The product of different chamograms, ~ the cross term described above, and the extended formula (22). That is, the line width of the pattern image is not only affected by the square of the components of each chamogram in formula (22). It is also affected by the cross-terms. As shown in Figure 22A and Figure 2: The line width distribution is distributed on the inclined ellipse by the combination of aberrations (Straight 22A The relationship between Z · 6 # Z.I3, the 22nd ugly picture is the relationship between ζ · 9 and m. These intersections like the S combination have sensitivity to the line width change. The 23rd figure shows the intersection Sensitivity (crosstalk) to the line width of the line pattern. In Equation (24), where S / 5jj is called, add the same value as in Equation ⑼. In Equation (24), if you only take For two terms, for example, it will be in the form of the following formula (25) .... (25) and formula can be seen, this shows Figure 22B The figure shows the oblique ellipse distribution. In fact, if the intersection term is expected to be calculated, then the existence of the intersection term can be confirmed in a considerable number of terms. From the π on X, it can be known that 'a certain aberration check CD-poly

The effect of the curve can be expressed by the change in the I line width of the focus direction of the CD_focus curve. Whichever is selected, in step 302 of the 24th, the simulation computer 46 calculates the projection optical system PL at each S measurement point by using an imaging simulator space image calculation in the state where there is actually aberration. CD-focus curve. The information on actual aberrations can also be deleted by the master control 50, and read from the memory device 42 for the wavefront aberrations obtained above. In step 304, the simulation computer 46 uses the above-mentioned step 90 200400540 302 'calculates the CD-focus curve using a fifth-order function shown in the following formula (26), and the formula is used to make the line through imaging simulation The difference function y, n of the difference is approximate. ) CD-focusing curve y'n = r 5η (χ — α n) 5+ 6 4η (χ— α, 2 Π 7 3η (χ — α η) 3+ 6 2 „(χ — α r 1 „(χ ~ α η) ... (26) In the formula, 7 5η, (J 4η, 7 3η, coefficients of each order of 2).

It can be combined with the number 7 5n, 5 4n, r 3n, Λ 0, * s ^ v 6 2n, and y in. It is also possible to use a linear combination that includes the > gram term and the c term. Specifically, the coefficients of the odd-numbered terms (hereinafter referred to as odd-numbered terms) 7 5,, e out s 1 nr 3n, r ln can be expressed by the linear combination of the components Cn, i of each violation term to show y ^ μ Table No, the coefficients of the even-order terms (the following are even-numbered terms) 6 4η, 5 2 are Noda Junshi η, you use the linear combination of the flat cn, i of each Zanek term component. Here, at -step 30 " ’the simulation computer 46 looks at the odd-numbered terms of y and n shown in (26), and # is calculated from the aerial image:

For the 5th, 3rd, and 1th time the Zernike terms of r ^ η and r 1η-s r 5i, S r 3i, s 7 ln. Figures 25 to 27 are tables respectively: Examples of the sensitivity Sr 51, Sr 3i, sr 1 " of each Zanike term. In the next step 308, the simulation computer 46 looks at the even term of the function 〆, and calculates the square of the components of each Zanek term through the aerial image calculation, the coefficient of the 4th order 5 4n, and the coefficient of the 2nd order $ 2n Sensitivity of each Zanike term is 8 5 4i, S 5 h. Figures 28 and 29 show the squares cn and i2 of each Chazlik item, respectively, as examples of the sensitivity s 5 4i, S (5 2i. 91 200400540 In step 301, the simulation is used by the computer village system. The following formulas (27) and (28) are used to find the coefficients of the fifth, third, and fifth of the measurement points n (n = 1 to 33) of the current aberration state of the projection optical system and the system PL.卩 n, &, τ 1 n, and 4th and 2nd-order coefficients (5 4n, (5 2n. 37 37 37 Y5n = iSy5, iCn ^ = ^ Sy3tiCn4 ^ = * · * (27) 37. 37 54w = 25 (54, A / δ2, = Σ5δ2, Α / ... (28) /: 1 i = l, with the above formula, all the functions l and (26) shown in the expression (23) can be calculated. The function y, n and the function y ,, n (= yn + y, n), and the CD-focus curve (also considering the deformer) of the measurement point n (= 1 to 33) are all predicted. Fig. 30 shows the representative CD-focus curves of the measurement points k and ^ (bu 丨 ~ 32) at each measurement point n (n = 1 to 33) under the set exposure conditions. y ', k, y, and k + 1. The pattern diagram at the time of k + 1. As shown in FIG. 30, the prediction method of this embodiment executes the above steps 2202 to 3301. With this, the 10-degree function y is used to approximate the CD-focus curve when the aberration of the projection optical system ρί is zero, and the linear combination of the components Ck, i of the search term k at the measurement point k is calculated to obtain The direction of the defocus amount (horizontal axis) and the line width (longitudinal axis) of the 10th-order function y are shifted by U ^ respectively, and further, the fifth-order function y'k is deformed to predict the amount. The function y, k of the measurement point k. As mentioned above, the wavefront on the pupil surface of the projection optical system PL varies depending on the measurement point, so the component of the Zernike term Cy also varies depending on the measurement point. Therefore, the above Y ”k, y” k + ι are different curves. 92 200400540 Figure 3 1Λ is an example of the CD-focus curve of measuring points 1, 1, 1, 17, 33 calculated by precise imaging simulation Figure 31B is an example of predicting the CD-focus curve of the measurement points i, 1 1, 17, and 33 under the same exposure conditions and the same pattern by using the aforementioned prediction method. As shown in Figures 31A and 31, The CD_focus curve simulated by imaging and the CD-focus curve predicted by the above-mentioned prediction method are very consistent at each measurement point, and can be highly accurate Measure the CD-focus curve. That is, if the above-mentioned prediction method is executed, the CD-focus curve can be predicted with high accuracy when the predetermined pattern is transferred under the predetermined exposure conditions. Figure 32 shows the line width deviation △ cd, the relationship between the calculation results using the ZS method described above and the method of directly calculating the spatial aberration by providing an appropriate wavefront aberration. Comparing this Fig. 32 with the above-mentioned Fig. 5 shows that if the new ZS method is used, the error can be reduced particularly. It can also be seen from Fig. 32 that the line width can be accurately calculated by the extended ZS method even if the aerial image is not performed by imaging simulation. In the above description of steps 202 to 301, the explanation is based on the premise that the operator or the like is used, but only the designation of the operator and the like is performed (step 202), and the processing after step 204 is performed on the computer 46 for simulation. (Or other computers connected to the proposed computer) Of course it can be done all. It is also possible to provide instructions from the host computer instead of the operations of the aforementioned operators. This deformation can be easily achieved by changing the software program. In a computer such as the intended computer 46, a processing program other than the above specified exposure conditions is performed. For example, it can be used on a CD (c0mpact disc), (digital versatile disc), M0 (magneto-optical disc), or 93. 200400540 Sales of information recording media such as FD (flexible disc). a π a Can be used as a steel outlet in the recording state. Of course, transfers are made through electronic content such as the Internet. The line can also be measured digitally (evaluation method). 2 Use the predicted measurement points as described above to measure the points. ^ Gather the lines and apply the evaluation of the exposure device 100. Add the image of the branch tea (or transfer image). ) Special feature 2: 1 As' As mentioned above, on the projection optical system ~ decent side, in

On / ... then, it is assumed that the positions of points 1 to η are provided with isolated line patterns. If the CD_focus curve is predicted on the right at each measurement point, then at each measurement point, according to the bias in the CD · focus curve. It can evaluate the characteristics of the isolated line pattern image in the exposure area U (for example, the in-plane uniformity).

Once again, as shown in Figure 5, suppose that a reticle ^ (which is arranged at a position corresponding to the measurement point Kn and arranged with mutually orthogonal line patterns) is executed for the vertical line ㈣ and the horizontal line pattern, respectively. If the above steps M2 to 3001 are used to predict the CD-focus curve, the CD_focus curve of the vertical line pattern and the horizontal line pattern shown in the aforementioned FIG. 8 is prepared, based on the difference of the CD_focus curve, It can evaluate the vertical and horizontal line width difference of each measurement point. _, That is, find the Zanek sensitivity to the change in the CD (line width) of each image of the vertical and horizontal lines, and find the Zanek term sensitivity S / at the intersection of the aberration (Zanek term) combination SmU Guan j), the combination of the symbols in the vertical and horizontal lines can find the combination of aberrations (Chaineck term) that affect the vh difference. The reason for this is to use the v-line and the h-line. Because of the same value except for the second term in equation (25), the difference between VH = △ CD (V) △ CD (H) is different. The vertical and horizontal lines are different when the sign of the parent and child's Zanek term sensitivity S / 5 i, j (i 妾 j) is different, the line width difference 94 200400540 is affected. In addition, suppose that the L / S pattern is arranged in the corresponding measurement points 1 to η. If the CD-focus curve is predicted for the line pattern image at both ends of the L / S pattern image, the CD- The difference in the focus curve can evaluate the difference in line width of the image at both ends of each measurement point, thereby, for example, the coma aberration of the projection optical system PL can be evaluated. In addition, if the above formulas (21), (22) (or (24)), (27), and (28) are combined and expressed in a matrix form, they are shown as the following formula (29)

Sa ^ a 33

A

Sa ^ Α, ι 37 " 33 A: Wa 0 0 0 0 0 0 7 ^ 33 0 Ψβ 0 0 0 0 0 A 0 0 Ψγ 0 0 0 0 = 0 0 0 Ψγ 0 0 0 y333 0 0 0 Ψγ 0 0 yii 0 0 0 0 0 ws 0 0 0 0 0 0 0 ws rhs Η 37 37,37 ^ 5,1 except 5,37 ^ 3,1 (29)

Outside 3,37 ^ 7ι, ι SYl, 37

SdA δ 4 δ 2, 33 '33

Sd, S5 4,37 2,1 S6 2,37 95 62 200400540 In the formula, W α, W stone, W r, w (5 series are respectively expressed as follows.

Cl, l… C1

Wa = Wy, 37 ai

Wj3 W3 C33, l… C33,37 ς, 22 c 1,37 C33, l2 ^ 33,22… Cu2 Cl /… c 1 p 2 3,1 C33,2 c 2 33.37 r 2 '1.37 2 33,37 • C '/ C33 / c C ^ 33,2 ^ 33,1 " 1,37 " 33,37

The matrix system differs because W does not choose either of equations (21) or (24). The formula (29) can be arranged as follows. 'f = Wa · ZS (30) In the formula, f is the matrix of the coefficients of the CD-focus curve, Wa is the matrix of wavefront aberrations, and ZS is the matrix of Zanek sensitivity. (Adjustment method) Secondly, according to the CD · focus curve predicted by the above-mentioned method of predicting the CD-focus curve, perform the evaluation method of the "transfer pattern" of the exposure equipment, and then use the evaluation result , Regarding the adjustment of the pattern transfer situation: the adjustment method will be explained. Here, the adjustment is aimed at improving the in-plane uniformity (the isolated line pattern image set in v step 10 02 described in the j). For example, even if all the patterns corresponding to each measurement point n (n = 1 ~ 33) are uniform, all of them are uniform. If the CD_focus curve between each measurement point deviates, the image will be imaged at that point. The pattern image is also not uniform. Therefore, the adjustment / method of this embodiment is as described above. In order to make the CD_focus curve of each measurement point as uniform as possible, the pattern transfer state of the exposure device 100 is adjusted. In the following, the This method will be explained first, according to the calculation formula of this adjustment method. In order to make the CD-focus curve of each measurement point 11 (n == 1 ~ 33) uniform, it is better to adjust the aforementioned 19 parameters to So that the above ^ 5η, 5 4n, r 3n, 5 2n, r ΐη, try to be as uniform as possible at each measurement point. Here, α η, η, 7, and η of each measurement point η (η = 1 ~ 33) In the same conditions between the measurement points, the target values of 4 7 3 η 5 2 η and γ 1 η have been used to make α η, / 5 n, r 5n, 5 4n, r 3n, d 2n, r ln tend to The adjustment amount of the adjustment parameters near each target value. As described above, when adjusting each adjustment parameter, the wavefront aberration of the projection optical system p L will change. If each matrix PARA1P, ~ pARAi9p is used to show the measurement point n (n = 1 to 33) of the aforementioned 19 adjustment parameters (taking these parameters as the adjustment parameters PARA1 to PARA19, respectively), each element of the fan matrix Wa above the adjustment amount per unit (that is, including ^^ 卜 丨 ~ "^ The change of the staff) ', for example, PARA 1P can be expressed by the following formula. MMIP '

Wa '0 0 0 0 0 0 · 0 that, 0 0 0 0 0 0 0 Wyy 0 0 0 0 0 0 0 Ψγ] 0 0 0 0 0 0 0 Wyf 0 0 0 0 0 0 0 WSf 0 0 0 0 0 0 0 WS} 97 (31) 200400540 In the expression, Wa, W,, wr, • W (5 'can use the following formula A, ... 7' ^ 1,37 Wa, = Wy ^ ·: ^ 33 , 1 ... 7 ^ 33,37 • VV… and 372. Heart Zu… Zπ2 · ▼ = * :: '··: ·· β .¾2 w ... z 2 ^ 33,37 · Z33，1 Z33,2Z33,1… Z33 372 「zu2 V… ZU72 · ㈣ 丨 = • • •. .¾2 Z33,22… Z33,372 ❿ In the formula, each matrix \ ¥«, \ ¥ / 3, \ ¥ 7 The Zn, i (i = l ~ 37) contained in each element of the circle (5,) is the change of C ′, i of the measurement point η of the adjustment parameter PARA1. Therefore, as shown in the following formula (32), if By multiplying the matrix vector PARAlPn 'on the column vector ZS (taking the sensitivity of the above coefficients as elements), the coefficient change amount (for example, α η) of the measurement point η per unit amount of the adjustment parameter PARA 1 can be obtained. (P1)) column vector B '[l]. (House 1) a (corpse 1) 33

A (corpse 1) β3 (ρ1) y33 Y51 (corpse 1) 98 200400540

5! [1] = PARAIP ^ ZS r5: (P1) r3, 33 (corporate 1) 7 ^ 33 r1! (House 1) (corporate 1) (corporal 1) 33δ4, (Ρ1) yi W1) (corporate 1) 02, (32)

W1) Similarly, for the adjustment parameters PARA2 to PARA19, the column vectors B '[2] to Β, [19] of the coefficient change amount of the measurement point η per unit quantity can also be obtained. Here, the vertical amount of each adjustment parameter adjustment amount is taken as the vertical amount P shown in the following formula (33). 'ADJX "

ADJ2 P =: (33) ADJ \ 9 ^ The change of the coefficient of the CD-focus curve to the adjustment amount of the adjustment parameters PARA1 to PARA19 Γ can be expressed by the following formula (34). / · = Mm [1] + dD / 2.5 丨 [2] +… ma 9 · 10,000, [19] (34) Here, the adjustment amount of the adjustment parameter and the CD-focus brought by the adjustment amount are used. The above relationship of the coefficients of the curve is calculated as shown below to make the pattern of each measurement point uniform. That is, let the column vector of each of the 2004-2004-4040 coefficient target values of the CD-focus curve be f, ^ ^ 〇 The column vector of each coefficient is f, by the column direction described above! B, "n ~ fvn Ω1 >", said moment 卩 A A 'formed by the first combination, these relations can be expressed by the following formula ⑽. "The matrix is ft-f = B · P (35) If the least square method is used to solve the problem ... If the above formula is solved, then it becomes the following formula (36). Ρ = (Β · Β) Ί〇 (36) In the formula, ΒΤ is the inverse matrix of the aforementioned matrix B (BT.B). V ΰ) ir 'This adjustment method is to calculate the column vector P of the adjustment amount by the main control device and calculate the formula (34) of 34. Find the difference ㈣1 ADil ~ Adjustment amount adJ19. In addition, in order to use the equation (36) to find the column vector p, the values of the coefficients of the CD-focus curve must be listed. The value (that is, the column vector ... But here, the car parent family As mentioned above, in order to improve the target of the isolated line pattern image, the "^" function of each measurement point is set to the same value. The target values of the coefficients of each order are all next, and the method of adjusting the vertical and horizontal line width difference As described above, 'Assume that the amount of the reticle h shown in Figure 5 should not be placed at the corresponding measurement points l ~ n. If the foregoing prediction method is performed, the same measurement will be performed. Point, the focus curve in ⑶. In each line pattern, this situation can be obtained. The same formula as the above formula (36) is also used, and the integer is fine to adjust The quantity is 9, but as mentioned above, in order to obtain a 2⑮CD-focus curve, the number of elements of the coefficient matrix ^ and f becomes the number of elements of f in the above formula (30) (33 > < 7 = 231 ) 2 times (ie ^ M) ,,, 100 200400540 If the target values of the vertical line pattern and horizontal line pattern coefficient corresponding to the same-measurement point are set to the same value, then the minimum width difference between the vertical and horizontal lines can be calculated. The adjustment amount ADJ1 to ADJ19. The second amount of the adjustment amount is set to 50. It is based on the memory ... Dm is the same as described above. The position and method of the movable lens are adjusted to y by the imaging performance i 48. At the same time as the above-mentioned actions, the master _ f, 1 # $.% Μ said ▲ The king & device 50 is to adjust the wavelength of the light source 16 according to the adjustment ϊ ADJ19. Snap button 7 to offset the lighting Light EL. In this embodiment, in the latter state, the adjustment of =: system, system ρ " can be further performed. Order the above steps 202 to 3 10 to predict the various measurements after adjustment. The point of the > gray-focus curve, further defrost the implementation of the above-mentioned evaluation method and adjustment status gradually approaching homogenization. The transfer of the measuring point pattern (exposure method) The exposure process line sheet R during the manufacture of semiconductor elements is a step-scan type exposure that is mounted on the marking line with Rs "for the components. Also, in this ~ ... hunting According to the actions described above, when performing a + β yoke-shaped exposure apparatus 100, the wafer w in the exposure area IA described above is calculated based on before calculation ~ ADJ18, $ > k W / placement and method The adjustment amount of Adj 16 ADJ18 is the same as that described above. In addition, in this embodiment, the actual transferred pattern is changed by exchanging the settings of the Japanese or Japanese, or the ruling, ruling, 21st, and 23h. As shown in Fig. 25 to Fig. 28, each check 101 200400540 Niek term component) changes, so to re-calculate these sensitivities, of course, it is necessary to re-execute the aforementioned prediction method, evaluation method, and adjustment method. As explained in detail above, if the above prediction method is adopted, then each aberration component cn obtained by expanding the wavefront aberration w (P, 0) series of the projection optical system pL according to the linear combination of the complex terms (respectively) , I (n == 1 ~ 33 '~ 37), can obtain the CD_focus curve (the variation curve of the pattern image projected through the projection optical system PL). Therefore, no complicated calculation that requires huge = calculation time is used The imaging simulation brought by, # 由 极 纯 二 运

(Find the linear combination of the terms Cni (n = 1 to 33, i = 1 to 37) of each aberration component). Under a given exposure condition, the projection optical system PL can be used to predict the relevant aberration status. The CD_focus curve of the pattern image can predict the characteristics of the projected image (or transfer image) of the pattern in a short time based on the prediction result. Also, if this prediction method is used, according to the movement of the CEU focus curve, the linear combination of the terms Cni (n == 1 ~ 33, i = 1 ~ 37) of each aberration component can also project the projection. From wavefront aberration w (U) of optical system PL

Because of the deformation of the CD-focus curve, the CD-focus curve can be predicted with higher accuracy. In addition, according to this prediction method, the direction of the image width axis direction c line width change direction CD. The movement of the focus curve is not only the square Cn / sensitivity of each aberration component, but also the cross terms of the aberration components that are different from each other. There is sensitivity too. If the linear combination of these intersection terms is further considered, the amount of movement in the direction of the image axis can be predicted with higher accuracy. W, the difference function V of the deformation of the variable n number of points n 1, the coefficient of the countable order term y sigh y η is shown in 102 200400540 wavefront aberration W of the projection optical system PL Among the aberration components Cn, i (n-1 ~ 33, 1 = 1 ~ 37) at (p, 0), since there is sensitivity, the linear combination of the components Cn, i of the Zernike terms can Coefficients for predicting odd order terms of the difference function. In addition, because the coefficients of the even-order terms of the difference function y, n have a sensitivity in the square C ^ 2 of the components of each Zanek term, it can be predicted by the linear combination of the squares Cn, i2 of the components of each Zanek term. The coefficients of the even-order terms can predict the deformation of the CD-focus curve in a short time and with high accuracy. In addition, if the above evaluation method is used, the above-mentioned prediction method is used, under a predetermined exposure condition, for each measurement point in the effective field of view of the projection optical system PL, it is possible to predict the transmission optical system in a short time and with high accuracy. The predetermined pattern image projected by the PL, therefore, based on the cd_focus curve, the characteristics (eg, uniformity) of the existing pattern image within the effective field of view of the projection optical system PL can be evaluated with high accuracy in a short time. Furthermore, if the adjustment method using this evaluation method is used, the evaluation method of this embodiment is used to evaluate the uniformity of a predetermined pattern image in the effective field of view of the projection optical system PL '&' According to the evaluation result, the transmission is adjusted Formation state of a predetermined pattern image of the projection optical system and the system PL. Therefore, according to the review

As a result, the characteristics of a predetermined pattern image can be adjusted to a desired state (for example, the direction of the uniformity of the printed image). T

The prediction method of the implementation mode of the favorable image is based on the assumption that the projection optical system PL selects a function of only 10 ... = curve of the even order term. Although the function is selected, the present invention is not limited to the south order of the °° function. The number may be less than 8 times, or more than 2 times. In any case, it is better that the function of the CD-focus curve is a high p; 103 200400540 even function. In the prediction method of the above embodiment, although the difference function 〆 is regarded as a 5th order function, the function may be below 4th order or may be above 6th order. Moreover, the prediction method of the above embodiment will correspond to each measurement. The pattern of the reticle for point measurement is used as a pattern of a vertical line pattern and a horizontal line pattern (ie, a cross pattern), or an isolated line pattern, but the present invention is not limited to this, and may be a plurality of parallel lines The pattern (L / S pattern) can also be a cross pattern: or a pattern combining parallel line patterns. Furthermore, it may include not only a vertical pattern: a horizontal pattern, but also a line pattern extending diagonally. In addition, when the [π pattern is used, the prediction method of this embodiment is used to predict the line widths of the line patterns at both ends of the L / s pattern, and the evaluation method of this embodiment is used to evaluate these line width differences (that is, the line "Wide outlier value" is the same as the adjustment method of this embodiment. According to the evaluation result, in the state where the pattern image is adjusted, the size of the line width outlier value is reduced, and the exposure can be performed with high precision if the "exposure is performed". In addition, although the evaluation method of the above-mentioned embodiment uses the in-sentence pattern, line width difference, and line width abnormal value in the line pattern plane as the evaluation items, the present invention is not limited to this, and can be based on the CD-focus curve. All items evaluated are regarded as evaluation items. Also, the above-mentioned evaluation method of the embodiment, although for each measurement point of the effective field of view of the projection optical system PL, predicts with a short time and high accuracy, the predetermined pattern image projected through the projection optical system pL under the predetermined exposure conditions. "CD · focus curve" Based on the CD · focus curve, comment on the situation of the projection optical system PL "Effective Uniformity of Vision" (2004, 200400540), and add the characteristics of the unpatterned pattern image (for example,

In one step, the wavefront aberration information about the information of the pattern projection image can be obtained. The sensitivity is based on the above-mentioned projection scene image (which is based on these data, and you can also consider chamunk, and expand the wavefront aberration series to check $ 克 = Chashenke polynomials uses arbitrary checks that affect the characteristics of the aforementioned projection image to change each other to evaluate the characteristics of the aforementioned pattern image. Even in the case of V type two items, the sensitivity of the test projection to the aforementioned projection image (which is a habit) = 相 = Characteristic of the second group of arbitrary grams == :: to evaluate the characteristics of the pattern image, so that the pattern image can be evaluated with higher accuracy. The evaluation method of the above embodiment is Belin, which is based on the use The optimal adjustment amount calculated in (36) above is not limited to those who automatically adjust by the imaging performance correction controller 48 and the like under the control of the main control device 50. Therefore, it can also be used according to the aforementioned adjustment amount. Manually adjust the imaging performance of the projection optical system, etc. In the prediction method of the above embodiment, various modifications can be considered. In addition, the above embodiment is explained by the following series of processes: according to the CD-focus curve A prediction method, an evaluation method for evaluating the pattern transfer state of the exposure device 100 using the predicted cd-focus curve, an adjustment method for adjusting the pattern transfer state based on the evaluation result, and an exposure method for performing exposure after the adjustment; A series of processes are used to perform all methods. The prediction method, evaluation method, and adjustment method of the present invention are implemented independently or in any combination 105 200400540. ^ Test. After continuing the prediction method of the above embodiment, the evaluation method, The adjustment method and the female exposure method are also implemented after various modifications of these prediction methods. In addition, after the implementation of the pre-method of the above embodiment and its modification, ^, phantom, except Applicable to the above-mentioned embodiment of the price method, adjustment method, exposure, soil + from above, especially the method, of course, can also be applied to the above various items as evaluation items of evaluation methods, manual adjustment methods, Various evaluation methods and exposure methods such as the exposure method using the step-and-repeat method. In 'on the projector used in terms of wavefront aberration measuring optical system PL wavefront aberration measurement, a shape may be used all having a replaceable wafer holder of the shape of the wavefront aberration measuring device, which

In this case, the wavefront aberration measurement is written. A group also moves the Japanese yen or wafer holder into the wafer table WST. From the wafer a, w σ WST, using the transfer system () Round loader, etc.), to perform automatic tearing of materials, such as door and order. In addition, in the above-mentioned embodiment, the wavefront aberration measuring device 1 can be freely attached and detached to the Japanese yen platform, and it can also be fixedly installed. At this time, only one part of the wavefront aberration measuring device 80 may be set on the wafer stage, and the other part may be arranged outside the wafer stage. Moreover, in the above embodiment, the wavefront aberration amount / 1 m 1 wind 4 is ignored. The light receiving light is used by the system as the aberration of the system, but the 5 L β wavefront aberration can also be used to determine the projection. Optical aberration, wavefront aberration. Also, in the wavefront aberration measurement and the measurement of the Ueda Keiuri, for example, when the aforementioned U.S. Patent No. 5 978 〇85 and the like are used, the measurement reticle for measurement is unrevealed by the temple. , For example, by exposing

The alignment system ALG provided by Youyizhi detects the position deviation of the reference pattern latent image of the pattern latent image transferred to the photoresist layer on the wafer. In addition, when the pattern of the latent image of the field test in double row is 106 200400540, a photoresist or the use of a dopant, a 7 ^ material, etc. can be used as a photosensitive layer on a substrate such as a wafer. With these methods, z- can perform the aforementioned adjustment of the projection optical system PL without the need for an operator or a service engineer through the exposure device 100 full self-reporting Wang Mudong. Also, in the aforementioned embodiment, “the optical element of the projection optical system pl is moved to adjust the imaging performance”, but it is not limited to this, and it can also be added to the driving mechanism: or instead of this mechanism, for example, the projection optical system PL is changed The mechanism of the gas pressure between the optical 7L pieces uses a mechanism that moves or tilts the reticle R to the optical axis direction of the projection optical system, or changes the optical thickness mechanism of the parallel plane plate arranged between the reticle and the crystal H # . However, the above embodiment uses 9 adjustment parameters, but the number and type of the adjustment parameters may be arbitrary, for example, the driving amount of the wafer surface (wafer platform wst) or the wavelength deviation of the illumination light EL may not be included. Shift etc. In addition, the above-mentioned embodiment has been described in the case of using a scanning exposure device as the exposure device, but is not limited to this. For example, a step-and-repeat type exposure device may be used. The application of the exposure device in this case is not limited to an exposure device for semiconductor manufacturing, for example, it can also be widely used in liquid crystal exposure devices that can transfer a pattern of a liquid crystal display element to an angular glass plate, and Display devices such as paste display or organic EL, exposure devices for photographic elements (Ccd, etc.), thin film magnetic heads, microcomputers, and DNA wafers. In addition, in order to manufacture micro-elements such as semiconductor elements, light exposure devices, EUV exposure devices, X-ray exposure devices, and electron line exposure devices, the present invention can also be applied to converting circuit patterns. Printed on glass substrates or silicon wafers and other exposure equipment. In addition, the light source of the exposure apparatus of the above embodiment is not limited to ultraviolet pulse light sources such as & laser, ArF excimer laser, KrF excimer laser, and a continuous light source can also be used. Ultra-high-pressure mercury lamps with radiant rays such as X-rays (wavelength: 365 nm). X-rays, especially EUV light, can also be used as the illumination light. Also, erbium-doped (or both europium and europium) can also be used. ) Fiber amplifier to amplify single-wavelength laser light in the infrared or visible range oscillated from DFB semiconductor laser or fiber laser, using nonlinear optical crystallization, the waveform is converted into high-order spectral waves of ultraviolet light. Also, projection optics The magnification ratio is not only a reduction system but also an equal magnification system and an enlargement system. Moreover, as far as the projection optical system is concerned, it is not limited to the refractive system, and a reflective refractive system having reflective optical elements and refractive optical elements, , Or the reflection system using only reflective optics. Also, when a reflective refraction system or a reflection system is used as the projection optical system PL, the aforementioned movable optics In terms of equipment, the position of reflective optical elements (concave mirrors, reflectors, etc.) is changed to adjust the imaging characteristics of the projection optical system. In terms of illumination light, especially when using Aq laser light or EUV light, etc. It is also possible to use the projection optical system PL as a total reflection system (consisting only of reflective optical elements). However, when using Ar, laser light, or EUV light, the reticle R can also be used as a reflection type. Also, When manufacturing the exposure device 100, etc., first, an illumination optical system i 2 including a plurality of lens elements, mirrors, and other optical elements is assembled as a single unit, and the projection optical system PL is regarded as a single 108 200400540. The system is composed of a reticle stage system and a wafer stage system made up of many mechanical components, and is assembled as a unit. In addition, optical performance is adjusted in order to achieve the desired performance of each unit. Mechanical adjustments, electrical adjustments, etc. When performing this adjustment, the projection optical system described in the above embodiment can be used for the projection optical system PL. The adjustment method, or the adjustment method of the pattern image characteristics of the projection optical system (including at least a part of the prediction method and the continuous evaluation method), is used for adjustment. Next, the illumination optical system 12 and the projection optical system PL are assembled. In the exposure apparatus main body, a reticle stage system and a wafer stage system are installed in the exposure apparatus main body to connect wiring and piping. Next, the illumination optical system 12 and the projection optical system pL are further optically adjusted. This is because these optical systems, especially the projection optical system pL, have subtle changes in imaging performance before and after being mounted on the exposure device. In this embodiment, the optical system of the projection optical system PL is performed after being assembled in the exposure device body. During the adjustment, the aforementioned wavefront aberration measurement device 80 is mounted on the wafer stage WST. The wavefront aberration is measured and measured in the same manner as described above, and the measurement result of the wavefront aberration is input to a computer. Steps, for example, calculate each lens element's 6-degree-of-freedom-direction adjustment ΐ 'to display the result of On the display. In accordance with this display, each lens element is adjusted by a technician (operator) or the like. With this, the adjustment of the projection optical system pL that does meet the desired imaging performance is completed. X. At this stage, because the aberrations that cannot be corrected can be judged, mainly high-order aberrations' are aberrations that are difficult to adjust automatically, so it is best to readjust the installation of the whole lens. If you can't get the performance you want, you must set up the optical two 1 ... In order to facilitate the projection optics in the exposure device: & H, the assembly of the projection optical system pl " Measure the wavefront aberrations using the dedicated wavefront volume "Setting etc." According to the measurement results, the presence and location of the optical element that needs to be reprocessed, etc., can also be adjusted at the same time. He Yuji

/ Then, further advance the comprehensive adjustment (electrical adjustment, operation confirmation, etc.), thereby making the exposure device such as the exposure device 10 of this embodiment, " The exposure exposure system can use the projection optical system PL which can adjust the optical characteristics with high precision , The pattern of the reticle R is accurately transferred on the wafer w. The manufacturing of the exposure device is preferably performed in a clean room with controlled temperature and cleanliness. (Element manufacturing method) Next, a component manufacturing method using the above-mentioned exposure apparatus for a lithography process will be described. Fig. 33 is a flowchart showing an example of manufacturing an element (a semiconductor wafer such as an IC or an LSI, a liquid crystal panel, a CCD, a thin film magnetic head, a microcomputer, etc.). As shown in FIG. 33, first, in step 401 (design step), the function and performance design of the device (for example, circuit design of a semiconductor device, etc.) is performed, and a pattern design for realizing the function is performed. Next, in step 402 (the step of manufacturing a light-shielding cover), a light-shielding cover is formed to form the designed circuit pattern. On the other hand, in step 403 (wafer manufacturing step), a material such as a silicon crystal is used to manufacture a wafer 0 to 200400540. In step 404 (wafer processing step), a step price of 403, 403 is used. The hood and wafer prepared by the householder will be described later. The actual circuit is formed on the wafer by micro-branch surgery. Secondly, in the step assembly step), the wafer processed in step 404 is used to perform the component & In step 405, if necessary, processes such as a dicing process, a bonding process, and a packaging process (wafer sealing) are included. AXON — Finally, in step 406 (check step ‘step’), perform step 405.

Check the operation confirmation test and durability test of each component. After the wearing, the components are completed and the components are shipped. King of clothing M. Fig. 34 shows a detailed flow example of the above-mentioned ㈣4 in a semiconductor device. In Fig. 34, in step 411 (oxidation step), both sides of the wafer are oxidized. In ㈣412 (CVD㈣), an insulating film is formed on the wafer surface. In step 413 (electrode formation step), electrodes are formed on the wafer by vapor deposition. In steps 4 and 4 (ion implantation step), ions are implanted into the wafer. Each of the above steps 411 to 414 constitutes each stage of wafer processing

Before the segment processing process, in each stage, as needed, optionally: Row 0 In each stage of the wafer process, if the above-mentioned pre-processing process is completed, the post-processing process is performed as described below. In this post-processing process, first, in step 415 (photoresist formation step), a photosensitizer is coated on the wafer. Next, in step 416 (exposure step), the circuit pattern of the hood is transferred onto the wafer by the exposure apparatus and exposure method described above. Next, in step 417 (developing step), the exposed wafer is developed 111 200400540 'In step 41 8 (etching step), the exposed members other than the photoresist residue are removed by the hunting worm I etch. . In addition, in the step 419 (photoresist removal step), the unnecessary photoresist is removed by repeating the pre-processing process and the post-processing process, thereby forming multiple circuits on the wafer. / If the device manufacturing method of the conventional embodiment described above is used, in the exposure process (step • step 416), the use of the exposure apparatus of the embodiment described above can effectively reduce the vertical line pattern and The horizontal line pattern transfer image has good line width difference between each other or isolated pattern line width uniformity. Therefore, the yield of the final product (element) can be improved, and the productivity can be improved. As explained above, The adjustment method of the projection optical system of the present invention is suitable for the adjustment of the projection optical system used for the projection of mutually orthogonal line pattern images. Furthermore, the prediction method and program of the present invention are suitable for predicting the transmission optical system The characteristics of the pattern image. In addition, the evaluation method of the present invention is suitable for the characteristics of the pattern image of a 5 flat ray through a projection optical system. The 5-week correction method of the present invention is suitable for adjustment The formation of the pattern image of the over-projection optical system is sad. 'The exposure method and exposure device of the present invention are suitable for transferring a pattern onto an object. Also, the method of manufacturing a component of the present invention is suitable for the production of a component. [Figure Brief description of the formula] (1) The first part of the drawing part is a schematic view showing the structure of an exposure device according to an embodiment of the present invention. 112 200400540 FIG. 2 is a cross-sectional view showing a wavefront aberration measuring device of a garden. Figure 3B shows the beam emitted from the mirror array when there is no aberration in the optical system. Figure 3B is a diagram showing the beam emitted from the micro-transparent | array during the aberration in the optical system. Method for adjusting the projection optical system PL for the purpose of line pattern image of orthogonal 2 axis direction ❼ 丨 the line width difference g 〇

Figure 5 is a plan view of the reticle for measurement viewed from the pattern surface side. Numbers 6A to 6F ® ′ are diagrams illustrating the wavefront disorder method of the pupil surface based on the changes in the values of the 9th and 12th terms of the Rick polynomial. Numbers 7A to 7F W are changes in the values of the 4th and 5th terms of the Chalkian polynomial based on the wavefront aberration name of the projection optical system. For example, the diagram of the wavefront disorder method of the pupil surface.

^ Figure 8 is a CD-focus used to explain the difference between the optimal focus position of the vertical line pattern and the horizontal line pattern, resulting in the line width difference between the vertical line pattern image (V) and the horizontal line pattern image (H). line graph. Fig. 9 shows the use of a wavelength of KrF laser with & Yun's as the light source 'illumination ρ = 0.75 · 2/3 # with lighting conditions, the numerical aperture (NL) of the projection optical system PL = 0 · At 680 hours, an example of the line-to-line difference (experimental result) of the vertical and horizontal lines obtained from the measurement of the line width of the photoresist image obtained from the graph on the reticle for transfer measurement. 113 200400540

40m part (part of the previous two paragraphs) diagram. zi2 = 40m λ, 20m λ ~ 20m X ^ —

Righteousness.

The contour diagrams of the image maps obtained by simulation. Figure 14 is a diagram showing an example of the calculation result of the zs (Zernike φ Sensitivity) of the line width deviation △ CD. Fig. 15 is a diagram showing the relationship between the calculation result using the conventional zs method and the calculation result using the aerial image regarding the line width deviation Δcd. Fig. 16 'is a flowchart showing a prediction method according to an embodiment (its Fig. 17A is an example of a} order function, and Fig. 7B is an example of the adjustment error. 苐 1 8 Figure 'shows an example of the sensation sensitivity sai. Figure 19 shows －50111 入 ～ 5〇111 入, using the distance of 10111 入 to calculate the amount of movement in the 11-point focus direction, using the minimum level Method to calculate an example of the calculation result of the straight line slope. Figure 20 shows the calculation result of the 11-point line width change using the same image calculation as in the case of Figure 19, assuming a second-order function, An example of the approximate result using the least square method. Fig. 21 is an example of the Zernike sensitivity S cold i. 114 200400540 Fig. 22A is a crosstalk diagram of z6 and Z13, and Fig. 22B is a diagram Crosstalk diagram of Z9 and Z12. Fig. 23 is an example of the sensitivity of each cross term. Fig. 2 4 'is a k-path diagram (part 2) showing a prediction method of an embodiment. Fig. 25 is a diagram showing An example of the sensitivity of the Zernike term sr 5i. Figure 26 shows the Zernike term. An example of the sensitivity of the sensitivity s r 3i. Fig. 27 is an example of the sensitivity of the s 7 π of the Zernike term.

Fig. 28 'is a graph showing an example of the sensitivity s 5 4i of the search term. FIG. 29 is a diagram showing an example of the sensitivity s 5 2i of the Zernike term. Fig. 30 is a diagram showing the operation when the CD-focus curve y ,, k, y''k + 1 is obtained. Figure 31A is an example of a CD-focus curve representing a representative measurement point by precise imaging simulation. Figure 31B is the same exposure condition predicted by a prediction method according to an embodiment of the present invention An example of the CD_focus curve of a representative measurement point with the same pattern.

Fig. 32 is a diagram showing the relationship between the calculation result of the new ZS method and the calculation result of the aerial image regarding Δ cd in the line width deviation. Fig. 33 is a flowchart showing an embodiment of a method for manufacturing a device according to the present invention. Figure κ34 is a detailed flowchart showing step 204 of Figure 33. (II) Symbols for element EL 1 Illumination light for exposure

IAR Illumination area 115 200400540 ΙΑ Illumination area (exposure area) LB Laser beam M Reflector PL Projection optical system R Graticule RST Graticule table TS Control information W Wafer WST Wafer table 11 Room 12 Illumination optical system 13 Refraction Optical elements 13i ~ 135 Lens element 15 Pupil aperture diaphragm 16 Light source 17 Light transmission window 20 Illuminance uniformity system 22 Optical integrator (Flying lens) 24 Illumination system aperture diaphragm 28A First relay lens 28A Second relay lens 30A Fixed reticle curtain 30B Moveable reticle curtain 32 Condensing lens

116 42 200 400 540 44 45 46 48 50 54R 54W 60a 60b 80 82 memory device display device input device computer imaging performance correction controller main control device reticle interferometer wafer interferometer irradiation system light receiving system wavefront aberration measurement device frame Body 82a Circular opening 84 Receiving optical system 84a Objective lens 84b Relay lens 84c Bending mirror 84d Collimator lens

84e 86 88 Micro lens array Light receiving section Glass cover Exposure device 117 100

Claims (1)

200400540 The scope of patent application: 1 method for adjusting the optical system of the shirt. The projection optical system projects the pattern image on the first surface onto the second surface. Contains information on the first optical characteristics of the projection optical system; + 2nd process of the second dimension, using the projection optical system, the first line pattern extending the first configuration in a predetermined direction, and the first line drawing The second line pattern image intersected by i is formed on the second surface, and the second line is measured: the line width of the first line pattern image and the second line width (the second line graph = hai line width) Difference; and <, Table I) The third process' is based on the difference between the property value obtained from the first! Process and the line width, to compare the optical characteristics of the first optical characteristics = system to control the cause and the science The size of the characteristic. The second light that affects the line width difference, the second, the middle, and the special = the adjustment method of the projection optical system of the first item in the range Aberration fund:; The information obtained in the process 1 is In the third process, the wavefront of the projection optical system is used to pay for the first Chagne coupon τ5 4 丨A skin order aberration 4th order: In the multiple burial items developed by any of the above sounds, when the size of the rotationally symmetric component term is not equal to the disaster, 4 Kang and the 2nd order rotationally symmetric component term " 'Root shadow optical system, to control the… 宽 wide difference' to adjust the size of the cast rotational symmetry component ^ 2nd order of the rotational symmetry component term is the same as the third order, such as the adjustment of the projection optical system of the second patent application scope Method 118 200400540 'In which, the 2nd-order rotationally symmetric component term is 4 12 terms, and the rotation collapse ^^ Eighth cos20 component term · The 9th term of the "term system 4th order" constituent term. 17 Patent scope The investment of the 2nd item, where the 2 ### Beixian method of adjustment of the system of the white% symmetry component term is the 4th order sin20 & ^ 13 item, the rotation ^ h ηη2θ component term of the second conversion The% component item is the 9th item of the 4th-order ㈣ component item. For example, the projection of the 2nd item of the patent application scope, where, the Chu 1 system. Fork ~ learn the adjustment method of the system first. Measure the $ # ^ plane of the projection optics to obtain information about the wavefront aberration. Technical ... For example, the second item in the patent application II encirclement, among which, the system of 坌 1, such as Wu Yu 、, and the method of death and death, is made by the Qin system, which is based on the complex array of the first surface, the dagger line pattern ... The difference between the two positions with different focusing positions _ the best setting when the image is formed? B ”Measured in each group, according to the measurement results, push ~ 2 Wave 7. The method of adjusting the projection optics in item 2 of the patent, please adjust the method 'h U 3rd process', in which the 2nd-order rotationally symmetric component term is large, 'zero' and measured in the 3rd process When the line width difference is non-zero, the size of the root symmetry component of the "Hai 2" and the line width difference adjust the projection light: the system 'adjusts the projection optics in such a way that the line width difference approaches the design value : System to optimize the size of the spin component term of the same order as the second-order rotationally symmetric component term of Yanhai. 8. The adjustment method of the projection optical system as described in the first item of the patent application scope, wherein the second process includes: / Image formation system Hunji 2 is formed from the object configured above. The image of a 2-line pattern; and. The line measurement process measures the first line width of the line width of the first line pattern image formed on the object, and the second line pattern formed on the object. The second line width of the image line width. 9. If the method of adjusting the projection optical system according to item 丨 of the patent application is based on the position control of the at least 1 degree of freedom of at least one fixed light-drying element constituting the projection optical system, and At least one of the pressure control of the gas in the light beam is used to control the magnitude of the second optical characteristic. · 10. If the adjustment method of the projection optical system of item 丨 of the patent application is "", the first line pattern is a vertical line pattern, and the second line pattern is a horizontal line pattern; "", the first optical The characteristics and the second optical characteristics are determined through the following processes: 1. A process for obtaining the Zernike sensitivity at the intersection of the Zernike term combination and the change in the line width of the vertical line pattern image and the horizontal line pattern image. ; And the process of finding the combination of Zanek's sensitivity symbols in the father ’s item in the vertical and horizontal lines. 11. The method of adjusting the projection optical system according to item 帛 of the patent application, wherein the information obtained in the first process is the information of wavefront aberration of the projection optical system; the first and second optical characteristics are obtained by using The polynomial polynomials are terms of the same order and different types among the plural polynomials of wavefront aberrations obtained by the first process in the series. 120 200400540 1 2. An exposure method, which uses a projection optical system to transfer the circuit pattern on the first surface to an object arranged on the second surface, which is characterized by including: adjusting the manufacturing process, using the scope of the patent application The method for adjusting the optical system of a shirt according to any one of the items ~ 1, to adjust the projection optical system; and the transfer process, which uses the adjusted projection optical system to transfer the circuit pattern onto the object. A type of exposure device is used to transfer the pattern formed on the hood to an object through the optical system for exposure. It is characterized by ... The adjusted projection optical system is used as the exposure optical system. 14. A kind of projection optical system * Yue Qinggu, i ϋ The ㈣ optics “is to project the pattern image on the 苐 1 plane onto the 8th and 隹 2 planes, which is characterized in that the first process of the package includes Information on the optical characteristics of the projection optics; and especially the 4 inch 14 second process, which is the value of the first property obtained according to the 1% of the first cloth, and the 4 inch set on the first surface. The difference between the line width of the line pattern extending in the predetermined direction and the second line orthogonal to the first ankle pattern of the younger one around the door. Adjust the line of the projection optics test wood to see the system to control the first 2. The second optical characteristic is formed by the projection optical system due to the size of the second optical characteristic. The line width and the second line pattern image line width :: Two 1-line pattern image 121 200400540 15. The adjustment method of the projection optical system such as item 14 of the scope of patent application, wherein the first line pattern is a vertical line pattern, and the second line pattern is a horizontal line pattern; 1 optical characteristics and 2 optical characteristics are obtained through the vertical line pattern image and the horizontal line The process of calculating the line width changes of the case image to find the Zanek sensitivity in the cross term of the Zanek term combination, and to find the Zanek terms of which the Zanek sensitivity symbols in the cross term are different in the vertical and horizontal lines. The combination process is determined. 16. An exposure method is to transfer the circuit pattern of Lu on the first surface to the object arranged on the second surface through a projection optical system, which is characterized by: including: adjusting the process, using the application The method of adjusting the projection optical system of item 14 or 15 of the patent scope to adjust the projection optical system; and the transfer process is to use the adjusted projection optical system to transfer the circuit pattern onto the object. An exposure device that transfers a pattern formed on a hood to an object through an exposure optical system, which is characterized by: · It has a projection optical system that is adjusted by using an adjustment method of a projection optical system according to item 14 of the patent application. , As the optical system for exposure. 〃 18. A method for adjusting the projection optical system, the projection optical system is a pattern on the first surface The projection on the second surface is characterized by: U, a process of obtaining wavefront aberration information of the projection optical system, a process of obtaining information about the pattern projection image, and 122 200400540 adjustment system, which is considered for use. The Zernike polynomial gives the wavefront aberration to a series of Zernike terms that are expanded in series. The interaction of any combination of Zernike terms that affects the characteristics of the projection image and the term on the projection image. To adjust the projection optical system according to the characteristic of the Zernike sensitivity. 19. If the method of adjusting the projection optical system of item 18 of the patent scope, the pattern includes a line pattern; the characteristic of the projection image includes the line The line width of the pattern. 20. An exposure method, which uses a projection optical system to transfer the circuit pattern on the first working surface to an object arranged on the second surface, which is characterized by _ in, including: adjustment process, which uses the first The method of adjusting the projection optical system of item 18 or item 19 to adjust the projection optical system; and a transfer process, which uses the adjusted projection optical system to transfer the circuit pattern onto the object. 21. An exposure device that transfers a pattern formed on a hood to an object through an optical system for exposure, which is characterized in that it is adjusted by a method of adjusting a projection optical system using the patent application No. 丨 8 A projection optical system is used as the exposure optical system. 22. A method for manufacturing an exposure device which transfers a pattern formed on a hood to an object through a projection optical system, which is characterized in that it includes the use of patent applications ranging from 1 to η, 14, 15, 18 The method of adjusting the projection optical system of any one of 19, 19 to adjust the manufacturing process of the projection optical system 123 200400540 system. 23. An exposure device that uses an energy beam to illuminate a pattern disposed on a first surface, and transfers the pattern to an object disposed on a second surface through a projection optical system, characterized in that: The measuring device in the living room is used to measure the optical characteristics including the first optical property of the projection optical system; the measuring device is used to measure the second optical surface formed by the projection optical system. The l-th line of the i-th plane extending in a predetermined direction The line width of the project image and the second line pattern image orthogonal to the first line pattern; the image formation state adjustment setting adjusts the formation state of the pattern image formed by the projection optical system; The liquid setting is based on the value of the first optical characteristic measured by the optical characteristic measuring device, and the ^ 1: the i-th line width of the pattern image line width and The second line pattern image line width of the second line = line width difference, the image formation state adjustment device to control the :: The magnitude of the optical characteristics that affects the line width difference due to the interaction of light to the light. , For example, the exposure device of the 23rd scope of the patent application, which includes a sales volume, a projection optics, and an aberration measurement device. Recursive aberration < Among them, the order above the wavefront image order is different from the order 2 difference. The exposure device of the 24th patent scope of the month has the characteristics of learning, which is the 2nd-order rotationally symmetric component term of the checksum gram term of wavefront aberration measured in the test set by series expansion using the test gram polynomial. The second optical characteristic, 124 200400540 Suspicious turn ;: The private component term is the same as the rotation-symmetric component term of the order. The female exposure device of the 25th patent scope, in the bean, the symmetry component item is 4 P both 2 Θ component of the 12th item and the 13th item, the rotational symmetry component item is 4th order 0 6 »component Item 9. For example, the exposure measurement wide measuring device of the 23rd scope of the patent application, the seven-in-line T ° Hai line Xia contains an aerial image measuring device for measuring the projection images of the images formed on the second surface. ^ 卞 2 g Λ.. Σ is the exposure device of the 23rd patent scope, which measures F 罟 ^ τ ′, and includes a photographing device that captures an image of the object fb formed on the second surface. . 29. The exposure device according to item 23 of the scope of patent application, wherein the image formation state adjustment I w # hall exhibition is a position adjustment in the direction of at least 1 degree of freedom of at least 1 optical element constituting the projection optical system , The adjustment of the gas pressure in a part of the optical path, the adjustment of the wavelength offset of the energy beam, and Related: At least one of the pattern-forming member formed by the pattern and the object ... At least one kind of adjustment is performed in the adjustment of the optical axis direction of the optical system. 30 types of component manufacturing methods include a lithography process, which is characterized in that: the lithography process uses an exposure device of any of the patent application scopes 13, 17, 21 to 29 for exposure. 3 1. A prediction method that predicts the characteristics of a pattern image formed through a projection optical system, which is characterized by including: ... obtained by performing series expansion using a predetermined formula based on wavefront aberrations of the projection optical system, respectively Contains the plural terms "linear combination 125 200400540" for each aberration, to calculate the variation curve (the displayed image size is relative to that after a given Λ 来 #, deviation from the best focus position < the second case is projected through the projection optical system According to the established map, the change curve is predicted based on the movement amount of the wavefront aberration caused by the movement amount of the wavefront aberration X Jin #. 32 If the prediction method of the 31st item in the scope of patent application, In the prediction system, the number of times i, the middle-order r ^ M ′ further includes: under the given exposure conditions, the operation of the helmet in the projection optical system and the projection of the optical system ... Μ to find the image The curve is a process curve in which the defocus amount changes, so that the obtained change curve approaches a higher-order function. · 33 'If the prediction method of the 32nd patent application range, # 中 ， 此 测 制' is based on the aberration components (the aberration components of the predetermined exposure conditions and the defocus amount Sensitivity as each system 'calculates the amount of movement of the variation curve in the direction of the defocus amount; ~ According to the "Haige aberration component (the square of the long aberration component under the given exposure conditions, the image size The sensitivity of the change is taken as the coefficient of each); the linear combination of the squares, # the amount of movement of the change curve in the direction of the change in the image size. , 34, "Please use the prediction method of patent range 帛 33. In #, the prediction process' in addition to the linear combination of the squares of the aberration components described above, it also uses the intersection terms (which are different from each other under the given exposure conditions). A linear combination of the aberration component's parent cross terms 'sensitivity to the direction of the image size change as each coefficient)' # The amount of movement of the change curve in the change of image size. Prediction method, in #, the higher-order function is a function composed of even-order terms only. Gate 126 200400540%, such as the prediction method of the 31st scope of the patent application:] system: identify complex terms that include each aberration component According to the linear junction, it is found that the variation of δH is caused by the movement of the maiden, the worker γ y y α 丄, and the wavefront aberration of Xiang Zhi, according to the amount of movement and the deformation, to predict the change curve. / κ 37. If the prediction method for the item% of the scope of the patent application, before the forecasting process, the advancement->-/ / This progress includes. The stage assumption assumes that the projection optical system has no image under the given exposure stop. Rates, pieces When the difference is high, the precision curve is used to find the small and fine warrior f Α Lu ~ the dynamic curve of the change in the defocus amount, so that the obtained variation curve approaches the process of the first-order function. 38. If the prediction method in item 37 of the patent application scope predicts the manufacturing process, enter _half-human M, ... ^ ^ 3 out of the process, the process is based on the actual aberration state of the projection 弁 wind i, not relevant The change curve of the pattern image after the projection;-. The prediction system is used to find the difference function (representing a higher-order function based on approximating the Γ motion curve of the shift amount 11), and expressing it using the calculation process. The difference of the variation curve function) is used as the variation of the variation curve due to the wavefront aberration. 39. For example, the prediction method of the% item in the scope of patent application, wherein the calculation process is performed in an analog manner. 40. For example, the prediction method of item 38 in the scope of patent application, wherein the pre-process is based on the aberration component of the aberration component based on the squared aberration function of the aberration component under the given exposure conditions. Even number The sensitivity of the order terms is as the coefficient of each) The square of the linear social person μ mountain # ^. S 'differs from the coefficients of the even-order terms of the difference function; 127 200400540 The prediction process is based on the aberration components (which will be in Under the given exposure conditions, the sensitivity of each aberration component to the odd-order terms of the difference function is used as the coefficient of each square) to calculate the coefficients of the odd-order terms of the difference function. The prediction method of item 31, wherein the existing gas system > 式 polynomial '50 a coefficient of each aberration component 查 圼 圼 term. 42. An evaluation method is to evaluate the pattern through the projection optical system The characteristics of the image are characterized by including a prediction process, which uses at least one measurement point within the effective field of view of the projection optical system, using the prediction method of any one of the 31st to 41st patent applications, Under the predetermined exposure conditions, through the projection optical system, the predicted change curve (represents the existing pattern image projected at the at least one measurement point, the size of the image is Defocus amount of change in bulk of the best focus position); and # private deduction system, the system according to the prediction result to evaluate the characteristics of the predetermined pattern image. g 8 43. The evaluation method according to item 42 of the scope of patent application, wherein the predetermined pattern is arranged corresponding to a plurality of measurement points in the effective field of view of the projection optical system; the side characteristics include the effective field of view of the projection optical system The homogeneity of the image. a 44. The evaluation method according to item 42 of the scope of patent application, wherein the predetermined pattern includes two line patterns, which are arranged in the direction of the optical axis of the projection optical system 128 200400540. An is orthogonal to each other; in the prediction process, the ^ Λ pattern predicts the variation curve. 々 system =, such as the evaluation method of the patent scope item 44, where the n 4 shell process' is the appraisal of the monk's achievements A ^ ”shell line pattern image line width difference between each other as the characteristics of the image. The evaluation method of 42 items, the fixed pattern of which includes /, dry, the existing well axis ^ It is set in a plane domain orthogonal to the anterior axis direction of the material science h parallel; Each line pattern predicts the variation curve. The method of evaluation of item 46 of the Ru Shenw patent scope, wherein the valence process is an evaluation result. The difference in line width between the line pattern images is used as the image characteristic. 48. An adjustment method is to adjust the formation state of the image through the projection optical system, which is characterized by the following: The present price process uses the evaluation method in the scope of patent application No. 42 to evaluate the corresponding projection optical system. Characteristics of a predetermined pattern image configured by at least i measurement points in an effective field of view; and The / adjustment process is to adjust the formation state of the predetermined pattern image transmitted through the projection optical system based on the evaluation result. I. 49. The adjustment method of item 48 in the patent scope, wherein the adjustment process uses the aberrations of the unit adjustment amount of the adjustment parameter about the measurement point to adjust the formation state of the predetermined pattern image. In the composition change, 'the sensitivity of each aberration component to the size of a given pattern image under the given exposure conditions' and the coefficients of each order term indicating the change curve of the size of the given pattern image to the defocusing movement. The deviation of the target value is used to calculate the adjustment amount of the adjustment parameter. Based on the calculated adjustment amount, the formation state of the predetermined pattern image of 129 200400540 is adjusted. · 50. The adjustment method according to item 49 of the scope of patent application, wherein the evaluation process is to evaluate the characteristics of a predetermined pattern image configured for a plurality of measurement points corresponding to the effective field of view of the projection optical system; The whole manufacturing process is to make the target value of the coefficient of the same term of the variation curve the same among the measurement points. 5 1. The adjustment method of item 49 in the scope of patent application, wherein, when a predetermined pattern includes a plurality of patterns, the target value of the coefficient of the same term of the variation curve is made the same among the patterns. _ 52. For the adjustment method of item 49 in the scope of patent application, the least square method is used to obtain the adjustment amount. 53. An exposure method is to transfer a circuit pattern on the i-th surface to an object arranged on the second surface through a projection optical system, which is characterized by including: adjusting the process, using the 43th scope of the patent application An adjustment method to adjust the formation state of the pattern image through the projection optical system; and a transfer process in which the circuit pattern is transferred through the projection optical system in the formation state of the adjusted image On the object. 5. A method for manufacturing a component, which includes a lithography process, which is characterized in that: the lithography process uses the exposure method of the 53rd aspect of the patent application. 55. An evaluation method for evaluating the characteristics of an image transmitted through a projection optical system, comprising: 130 200400540 obtaining the wavefront aberration m range of the projection optics; a process of obtaining projection image information about the pattern; and / Ping ^, manufacturing process, considering the Chanck's sensitivity to the projection image (by using the Zanxian term to expand the wavefront aberration by a series of multiple Zanek terms, the interaction of the projection image characteristics The characteristic change of the arbitrary Zanek term and the parent term of the mouth) that affects the pattern image characteristics. 56. For example, the evaluation method of the 55th aspect of the patent application scope, wherein the pattern includes a line pattern, The characteristics of the projection image include the line width of the line pattern. This method is to adjust the formation state of the pattern image through the projection optical system, and is characterized by including an evaluation process that uses the evaluation of the 55th scope of the patent application. Method, the characteristics of a predetermined pattern image corresponding to at least one measurement point in the effective field of view of the projection optical system; and the whole process, which is Based on the evaluation results, the formation state of the predetermined pattern image through the projection system is adjusted. 58. An exposure method in which a pattern on the first surface is transferred to a second surface through a projection optical system The object is characterized in that it includes: the adjustment process is to adjust the formation state of the pattern image through the projection optical system using the adjustment method of item 57 of the patent application scope; and the transfer process is after the adjustment In the image forming state, the pattern is transferred to the object through the projection optical system. 59. A component manufacturing method includes a lithography process, which is characterized by: 131 200400540 The lithography process uses an application patent The exposure method in the range of item 58. 60. A recording medium for recording the prediction of the characteristics of a pattern image through a projection optical system for recording by a computer, which is characterized by: ° This program causes the computer to execute the following procedures, that is, Complex numbers of aberration components obtained by expanding the wavefront aberration series of the projection optical system using a predetermined formula The linear combination is used to show the amount of movement (related to: the image of a predetermined pattern projected through the projection optical system under a predetermined exposure condition, and the change in the size of the image relative to the defocus amount from the optimal focus position is displayed. The curve is caused by the wavefront aberration), and the prediction procedure of the change curve is predicted based on the calculated movement amount. 61. For example, the record of the 60th item in the scope of patent application ^ Y, the formula is before the prediction step. , Further causing the computer to execute the following procedure, that is, assuming that the projection optical system helmet image is required for 2 days under the given exposure condition, to make the change in the size of the image showing the change in the defocus amount centered on the surface level Function program 62. For example, the recording medium of the 61st patent application scope, of which $. The formula causes a computer to execute the following program as the prediction program. According to the respective aberration components θ ",, where The combination of Wang, Wang, and σ is used to predict the procedure of the amount of movement of the variation curve in the direction of the defocus amount. According to the aberration components that will change the size of the image under the given exposure conditions, Sensitivity as a coefficient of each aberration component square ^ linear = combined to predict the program about the change of the change curve in the image size °.-The amount of movement in the direction of 132 132005005 ... 63, such as the scope of the patent application No. 62 The recording medium, "The program, makes the computer execute the following 庠 as the material, that is, in addition to the linear combination of the previous '77, 'squares, but also according to the cross terms (passing the predetermined exposure conditions to each other) The procedure of the intersection of the different aberration components with each other (= the sensitivity of the direction of change in the size of the image as each coefficient) is a procedure related to the amount of movement of the change curve in the change in the size of the image. Recording media of 61 items, in which the order function is composed only of even-order items < ^. / 65 As a recording medium of 60 items that declares a patent scope, ^ is to make the order of the electric line, Incoming material: that is, = the linear curve of the plural terms that do not contain the aberration components is caused by the scab rib of the wave surface aberration, ff is derived, the interaction μ, the distortion of the aberration, according to the The amount of movement and the deformation condition to predict the change curve. Ai ... 66: If the recording medium of the scope of patent application No. 65, where the program is before the Dhai pre-private sequence, the computer is further executed The following procedure, that is, assuming that the projection optical system has no aberrations under the given exposure conditions, makes the size of the displayed image defocused to a higher-order function. The variation curve of the interaction approaches 67 The recording medium applying for 66 items in the patent scope of the month, in the basin, is: before the prediction process, the computer is further executed :, under two: light conditions, the projection optical system through the actual aberration state: Regarding a predetermined pattern image after projection, the Only when the program is changed, the prediction of the change in the amount of political focus is used as the prediction program, so that the computer executes the difference between the higher-order function (moved according to the shift 133 200400540 momentum) and the change function (obtained by the above calculation program). The function is used as a program resulting from the variation of the wave curve of the wavefront aberration. 68. The recording medium such as the 67th in the scope of patent application, wherein the program causes the computer to execute the following procedure to make the prediction procedure. :, Predict the difference based on the linear combination of the squares of the aberration components (the sensitivity of the squared aberrations of the aberrations to the even-order terms of the difference function under the given exposure conditions as the coefficients of each) Coefficients of even-order terms Based on the silly scepter, 7/8 / Guangcheng injury (the sensitivity of the countable order term of the difference of six t number of the letter age under the given exposure conditions is taken as the coefficient of each), , 'Combined to predict the coefficients of the odd-order terms of the difference function, the application is called the recording medium of any one of the 60th to 68th of the patent scope', where the established formula # 杏 系 查 耶克 polynomial, the Each aberration component is a coefficient of each cubic term. ^ Pick up, schema: as next page 134