TW200931290A - Mask data generation method, mask fabrication method, exposure method, device fabrication method, and storage medium - Google Patents

Abstract

The invention provides a generation method of generating data of a mask, comprising a calculation step of calculating an aerial image formed on an image plane of a projection optical system, an extraction step of extracting a two-dimensional image from the aerial image, a determination step of determining a main pattern of the mask based on the two-dimensional image, an extraction step of extracting, from the aerial image, a peak portion at which a light intensity takes a peak value in a region other than a region in which the main pattern is projected, a determination step of determining an assist pattern based on the light intensity of the peak portion, and a generation stop of inserting the assist pattern into a portion of the mask, which corresponds to the peak portion, thereby generating, as the data of the mask, pattern data including the assist pattern and the main pattern.

Description

200931290 IX. Description of the Invention [Technical Field] The present invention relates to a mask data generating method, a mask manufacturing method, an exposure method, a device manufacturing method, and a storage medium. [Prior Art] Description of Related Art ◎ The exposure apparatus uses photolithography to fabricate a micropatterned semiconductor device such as a semiconductor memory or a logic circuit. The exposure device projects and transfers the circuit pattern formed on the mask (mask) onto a substrate (e.g., a wafer) by a projection optical system. With the recent advancement in micropatterning of semiconductor devices, the pattern formed by the exposure apparatus must have a line width smaller than the exposure wavelength (the wavelength of the exposed light). However, in such a micropattern, light diffraction may occur, so the pattern formed on the wafer (the shape of the pattern) is not complete enough, for example, the pattern corner becomes 0 circle, or the pattern length becomes short. In recent years, in order to reduce the accuracy of pattern shape formed on a wafer, a pattern shape correction process (so-called optical proximity correction, Optical Proximity Correction, 0PC) is used to design a mask pattern. 〇PC correction is based on a rule-based system or a model-based system. Light simulation is used to calculate the influence of the shape of each component of the mask pattern and its surrounding components to correct the shape of the pattern. . The pattern base system uses light simulation to deform the mask pattern until the desired light image is obtained. A method of inserting a helper feature that is too small to be decomposed is also suggested. The technique for calculating how to insert an auxiliary pattern by borrowing 发表 is published as follows: Japanese Patent Publication No. 2004-221594 (Patent Reference 1), and Robert Socha, Douglas Van Den Broeke, Stephen Hsu, J. Fung Chen, Tom Laidig, Noel Corcoran, and Will Conley, "Contact Hole Reticle Optimization by Using Interference Mapping Lithoraphy (IML (TM))" by Proceedings of SPIE, USA, SPIE press, 2005, vol. 5838, PP 180~193.値 Calculate the interferogram, thus obtaining the position that causes interference on the mask, and the position where the interference is offset on the mask. The interference pattern is inserted on the interferogram, and the auxiliary pattern is inserted so that the pass will be transferred. The phase of the light of the exposure of the main pattern is equal to the phase of the light exposed by the auxiliary pattern. At the offset position on the interferogram, an auxiliary pattern is inserted so that the phase of the light passing through the exposure of the main pattern is passed. The optical phases of the exposure pattern of the auxiliary pattern are different from each other by 180 degrees. As a result, the main pattern to be transferred and the auxiliary pattern are strongly interfered with each other, so that the master can be successfully The above-mentioned interferogram represents the light amplitude on the image plane, and the position of the image plane has an imaging relation with the mask plane. The main pattern is the component existing on the mask and will be Transfer to the wafer. The circuit pattern can be roughly divided into line pattern and contact hole pattern. Patent Reference 1 The assumption of calculating the auxiliary pattern is equal to the one-dimensional space line, and the contact hole pattern is equal to the space-free Point; therefore, the shape of the main pattern cannot be calculated. To overcome this situation, after this calculation, the main pattern must be recalculated, such as the position, shape, and size of the auxiliary pattern. - 200931290 The general rule is: not to calculate the specification of the aerial image, but to calculate the specification from the non-approximate aerial image based on the model basic system. In view of this, the patent reference 1 must calculate the non-approximate aerial image many times to obtain the main image. The mask pattern of the type and the auxiliary pattern requires a long calculation time. Moreover, since the patent reference 1 calculates the auxiliary pattern, the main pattern is assumed by line or point. Therefore, the interaction between the light approximation effect between the main pattern and the auxiliary pattern cannot be accurately estimated. The subsequent corrected main pattern will first obtain the auxiliary pattern of φ, resulting in an optical approximation effect. Therefore, the auxiliary map may not be obtained. The type of predictive effect, or the auxiliary pattern, may have the opposite effect on the resulting mask pattern. In particular, when the line pattern is used as the main pattern, the auxiliary pattern is difficult to insert because the light approximation is greatly changed in shape due to the edge portion and the curved portion of the line. SUMMARY OF THE INVENTION The present invention provides a method of generating mask data, which forms a highly accurate micropattern. According to a first aspect of the present invention, there is provided a method of generating a mask material for use in an exposure apparatus by a computer; the exposure apparatus comprising: an illumination optical system for illuminating the mask with light from a light source, and a pattern of the mask a projection optical system projected on a substrate; the method comprising: an aerial image (aeria 1 image) calculation step, according to a light intensity distribution formed on a pupil plane of the projection optical system Information relating to the wavelength of the light of the light source, information relating to a numerical aperture on one side of the image plane of the projection -6-200931290 optical system, and a target to be formed on the substrate a pattern for calculating an aerial image formed on an image plane of the projection optical system; a two-dimensional image extraction step of extracting a two-dimensional image from the aerial image calculated in the aerial image calculation step; Step, according to the two-dimensional image extracted by the two-dimensional image extraction step, determining a main pattern of the mask; a peak portion Extracting step, extracting a peak portion from the aerial image calculated by the aerial image calculation step; the peak portion is in addition to the region where the main pattern is projected on the image plane, and the light intensity is at a peak; a pattern determining step of determining an auxiliary pattern according to the light intensity of the peak portion extracted by the peak portion extraction step; and a generating step of inserting the auxiliary pattern determined by the auxiliary pattern determining step into the mask a portion of the cover corresponding to the peak portion extracted in the peak portion extraction step; thereby generating pattern data including the auxiliary pattern and the main pattern determined in the main pattern determining step, The material for the mask. Q According to a second aspect of the present invention, there is provided a method of generating a mask material for use in an exposure apparatus by a computer; the exposure apparatus comprising: an illumination optical system for illuminating the mask with light from a light source, and a mask a projection optical system projected on a substrate; the method comprising: a first aerial image calculation step, according to information related to light intensity distribution formed on a pupil plane of the projection optical system, and light from the light source Wavelength-related information, information related to a digital aperture on one side of the image plane of the projection optical system, and a target pattern formed on the substrate, and calculating an aerial image formed on an image plane of the projection optical system a first 200931290 two-dimensional image extraction step of extracting a two-dimensional image from the aerial image calculated in the first aerial image calculation step; a first main pattern determining step, according to the first two-dimensional image extraction step Extracting the two-dimensional image to determine a main pattern of the mask; a first peak portion extraction step from the first aerial image meter Calculating a peak portion of the aerial image calculated in the step; the peak portion is in a peak region except that the main pattern is projected on the image plane; the first auxiliary pattern determining step is performed according to the first a peak portion is extracted by the light intensity of the peak portion extracted by the step, and an auxiliary pattern is determined; the second aerial image calculation step is performed according to a pattern and the aperture of the illumination optical system in the projection optical system. Calculating information related to the light intensity distribution formed on the plane, information related to the wavelength of the light of the light source, and information related to the digital aperture on the image plane side of the projection optical system, and calculating the image in the projection optical system An aerial image formed on a plane, the pattern including the main pattern determined in the first main pattern determining step and the auxiliary pattern determined in the first auxiliary pattern determining step, the auxiliary pattern φ And inserting a portion of the mask corresponding to the peak portion extracted in the first peak portion extraction step; a second two-dimensional image extraction step, calculating a step from the second aerial image Extracting the two-dimensional image from the aerial image calculated in the step; the second main pattern determining step determines the main pattern of the mask according to the two-dimensional image extracted by the second two-dimensional image extraction step; a peak portion extraction step of extracting a peak portion from the aerial image calculated in the second aerial image calculation step; the peak portion is in a peak intensity except that the main pattern is projected on the image plane a second auxiliary pattern determining step of determining an auxiliary pattern according to the light intensity of the peak portion extracted by the second peak portion extraction step -8-200931290; a generating step, the second auxiliary pattern determining step Determining the auxiliary pattern into a portion of the mask, the portion corresponding to the peak portion extracted in the second peak portion extraction step; thereby generating the auxiliary pattern and the second main pattern Determining the pattern data of the main pattern determined in the step as the material of the mask. According to a third aspect of the present invention, a mask manufacturing method is provided, comprising: fabricating a mask in accordance with data generated by the foregoing generation method. According to a fourth aspect of the present invention, there is provided an exposure method comprising the steps of: fabricating a mask by the mask manufacturing method; irradiating the manufactured mask; and forming the mask with the projection optical system The type of image is projected onto a substrate. According to a fifth aspect of the present invention, there is provided a device manufacturing method comprising the steps of: exposing a substrate using the exposure method described above; and performing a development process on the exposed substrate. According to the sixth aspect of the present invention, a storage medium is provided, and the storage method Q causes the computer to generate mask data for use by the exposure device; the exposure device includes: an illumination optical system that illuminates the mask with light of the light source, and The pattern of the mask is projected onto a projection optical system on the substrate; the medium causes the computer to perform the following steps: an aerial image calculation step, according to the light intensity distribution formed by the illumination optical system on the pupil plane of the projection optical system Information, information about the wavelength of the light of the light source, information on a digital aperture on one side of the image plane of the projection optical system, and a target pattern formed on the substrate, calculated in the projection optical system An aerial image formed on the image plane; a two-dimensional image extraction step of extracting the two-dimensional image from the aerial image calculated in the aerial image calculation -9-200931290; a main pattern determining step according to the two-dimensional image The two-dimensional image extracted by the image extraction step determines the main pattern of the mask; a peak portion extraction step from the aerial image The aerial image calculated in the calculating step extracts a peak portion; the peak portion is in the peak region except for the region where the main pattern is projected on the image plane; an auxiliary pattern determining step is performed according to the peak portion Extracting the light intensity of the peak portion extracted by the step, determining an auxiliary pattern; and a step of producing the auxiliary pattern, the auxiliary pattern determined by the auxiliary pattern determining step is inserted into a part of the mask, the portion Corresponding to the peak portion extracted in the peak portion extraction step; thereby generating pattern data including the auxiliary pattern and the main pattern determined in the main pattern determining step as the material of the mask. According to a seventh aspect of the present invention, a storage medium is provided, and the stored program causes the computer to generate mask data for use by the exposure device; the exposure device includes: an illumination optical system that illuminates the mask with light of the light source, and the mask a projection optical system projected onto the substrate; the medium causes the computer to perform Q: the first aerial image calculation step is based on the light intensity distribution formed by the illumination optical system on the pupil plane of the projection optical system Information, information about the wavelength of the light of the light source, information on a digital aperture on one side of the image plane of the projection optical system, and a target pattern formed on the substrate, calculated in the projection optical system An aerial image formed on the image plane; a first two-dimensional image extraction step of extracting the two-dimensional image from the aerial image calculated in the first aerial image calculation step; the first main pattern determining step, according to the first The two-dimensional image extracted by the two-dimensional image extraction step determines the main pattern of the mask; the first peak portion is extracted step-10-200931290 Extracting a peak portion from the aerial image calculated in the first aerial image calculation step; the peak portion is in a peak region except for the region where the main pattern is projected on the image plane; the first auxiliary map a determining step of determining an auxiliary pattern according to the light intensity of the peak portion extracted by the first peak portion extraction step; and a second aerial image calculating step according to a pattern and the projection optical system at the projection optical The information related to the light intensity distribution formed on the pupil plane of the system, the Q information related to the wavelength of the light of the light source, and the information related to the digital aperture on the image plane side of the projection optical system are calculated in the projection An aerial image formed on the image plane of the optical system; the pattern includes a main pattern determined in the first main pattern determining step and an auxiliary pattern determined in the first auxiliary pattern determining step, The auxiliary pattern is inserted into a portion of the mask corresponding to the peak portion extracted in the first peak portion extraction step; the second two-dimensional image extraction step is from the first Extracting the two-dimensional image from the aerial image calculated in the aerial image calculation step; the second main pattern determining step determines the main image of the mask according to the two-dimensional image extracted by the second two-dimensional image extraction step a first peak portion extraction step of extracting a peak portion from the aerial image calculated in the second aerial image calculation step; the peak portion is a light intensity system other than the region in which the main pattern is projected on the image plane In the second auxiliary pattern determining step, determining an auxiliary pattern according to the light intensity of the peak portion extracted by the extraction step of the peak portion; and a generating step, determining the second auxiliary pattern determining step Inserting an auxiliary pattern into a portion of the mask, the portion corresponding to the peak portion extracted in the second peak portion extraction step; thereby generating the auxiliary pattern and determining in the second main pattern determining step -11 - 200931290 The pattern data of the main pattern is used as the material of the mask. Further features of the present invention will become apparent from the following description of the embodiments of the drawing. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS In the following, the preferred embodiments of the present invention will be described with reference to the accompanying drawings, The present invention is applied to a mask-generating material (mask pattern) for use in micromechanics and in manufacturing various devices such as an integrated circuit (1C) and a large integrated circuit (LSI). A semiconductor wafer, a display device such as a liquid crystal panel, a detecting device such as a magnetic head, and an image sensing device such as a charge coupled device (CCD). Micromechanics here refers to the manufacturing technology of micro-precision mechanical systems using micro-structured semiconductor circuit manufacturing technology, or the mechanical system itself. For example, the present invention is suitable for generating mask data (mask pattern) for use in an exposure apparatus having a high numerical aperture (NA) of its projection optical system, and filling a liquid gap between its projection optical system and the wafer. Immersion exposure apparatus ° The concepts disclosed in the present invention can be mathematically modeled. Thus, the present invention can be implemented with software functions within a computer system. The software functions in the computer system include programs with executable software codecs and generate mask data that can form high-definition micro-patterns. The software codec is executed by a processor of a computer system. At -12-

❹ 200931290 During the software coding operation, the codec or related data is in the platform; however, the 'software codec is often stored in the appropriate computer system. The software codec can be set as a group, and is placed on at least one recording medium for reading by the computer, and is described by the foregoing codec. It can be one or several schematic diagrams of FIG. 1 to show the execution basis of the processing device 1. One of the methods of the present invention is directed to a method of production. This produces mask data. The processing device 1 can be shown as a general computer, comprising: a bus line 10, a control unit 20, a memory unit 40, an input unit 50, and a media interface 60. The bus line 1 interconnects the control unit 20, the display unit 40, the input unit 50, and the media interface 60. The control unit 20 is controlled by a central processing unit (CPU), a digital signal processor (DSP), or a micro memory and includes a cache memory for temporarily inputting the mask data generating program 401 by the input unit 50. 20 is started up and executed in the storage slip generation program 401. The control unit 20 performs a method related to the mask data generation using the storage sheet 7 (the processing will be described later. The display unit 30 may be constituted by a CRT display or a liquid crystal display device; the display unit 30 may display information related to the execution program 401 (such as the latter) The near two-dimensional image 410 and the mask data 408 will be described. The ί recording is stored in the computer 'other locations or 'made as one or several modules.' The software product of the present invention. I structure, this processing device, The production method is used for I; and as shown in Figure 1, the unit 30, the storage unit, the storage unit, the graphics processor computer, etc., and the user's startup command, Ϊ; 40 masking capital 40 The mask data of the mathematics indicator or the like of the present invention generates an aerial image-like 200931290. The storage unit 40 may be constituted by a memory or a hard disk; the mask data generating program 401 stored by the storage unit 40 is derived from the media interface 60. Connected storage medium 70. The storage unit 40 stores the graphic data 402, the effective light source information 403, the NA information 404, the information 405, the aberration information 406, and the resist information 407 for performing the mask data generation program. 40 1 hour input information set. The storage unit 40 stores the mask data 408 for outputting information after the mask data generating program 40 1 is executed. The storage unit 40 also stores the approximate aerial image 409, the two-dimensional image 410 of the approximate aerial image, and the deformed graphic data (the main pattern and the auxiliary (subsidy) pattern) 411 as the execution of the mask data generating program 40 1 . Temporary data set. The mask data generation program 401 is used to generate the mask data 40 8 to indicate the mask pattern used by the exposure device or the pattern formed by the pattern forming unit of the spatial light modulator (SLM). Graphic type. In this case, the graphic elements are composed of polygons, and a set of polygons constitutes the entire mask pattern. Pattern data 402 is information about the pattern design, for example, designing an integrated circuit (the necessary pattern to be formed on the wafer, which will be referred to as a design pattern or a target pattern later). The effective light source information 403 is about polarization and light intensity distribution (effective light source) formed on the pupil plane of the projection optical system of the exposure apparatus. The NA information 404 is about the digital aperture (ΝΑ) on the image plane side of the -14- 200931290 projection optical system of the exposure device. The information 40 5 is the wavelength of light (light for exposure) emitted by the light source of the exposure device. The aberration information 406 is related to the aberration of the projection optical system of the exposure apparatus. 〇 The photoresist information 407 relates to the photoresist applied to the wafer. The mask data 408 represents a pattern of a real mask generated by the mask data generation program 401. The approximate aerial image 409 is a distribution of approximate aerial images formed by the interference between the main diffracted beams representing the surface of the wafer during the execution of the mask data generation program 401. The two-dimensional image 4 1 0 is generated during the execution of the mask data generation program 410, and is equal to the two-dimensional image obtained by cutting the approximate aerial image 409 according to the slice value. The deformed pattern data 411 includes a main pattern deformed by executing the mask data generating program Q 401 and an auxiliary pattern inserted by executing the mask data generating program 401. The pattern data 402, the mask data 408, and the deformed pattern data 411 include information sets such as position, size, shape, transmittance, and phase of the main pattern and the auxiliary pattern. The pattern data 4〇2, the mask data 40 8 and the deformed pattern data 411 also include information sets such as transmittance and phase of the (background) region where the main pattern and the auxiliary pattern are not present. The input unit 50 includes: a keyboard, a mouse, and the like. The user can input the input information set such as the mask data generation program 401 by inputting -15-200931290 unit 50. The media interface includes, for example, a floppy disc driver, a CD-ROM driver, a USB interface, and the like, and can be connected to the storage medium 70. The storage medium 70 includes, for example, a floppy disk, a CD-ROM, a USB memory, etc., and provides a mask data generating program 401 and other programs executed by the processing device 1. 0 Next, a flow of generating a mask data by the control unit 20 of the processing device 1 executing a mask data generating program will be described with reference to FIG. The control unit 20 calculates an approximate aerial image (ie, an aerial image of the target image) according to an information set (pattern data, effective light source information, NA information, lambda information, aberration information, and photoresist information) in step S102'. . Please note that the input information set (pattern data, effective light source information, Α information, λ information, aberration information, and photoresist information) is input by the user via the input unit 50 and stored in the storage unit 40. There are two reasons why φ calculates an approximate aerial image in step s 1 02 instead of calculating a precise aerial image. First, the calculation of an approximate aerial image is much shorter than the time taken to calculate a precise aerial image; second, the approximate aerial image enhancement pattern coherency' thus clearly shows the light approximation. Various approximate aerial image calculation methods have been published, for example, those published in Patent Reference 1, and approximate aerial images can be calculated by interferogram deformation. In this case, a singular value decomposition is performed on the TCC (transmission cross coefficient). Let λ; be the ith eigenvalue, and d>i(f,g) be the ith eigenfunction -16- 200931290 (eigenfunction). Note that (f, g) is the coordinate on the pupil plane of the projection optics; TCC indicates the coherence of the effective source (the degree of homology depending on the distance of the mask surface). According to Patent Reference 1, the dry graph e (x, y) is the sum of several eigenfunctions, expressed as follows: N, e(x, y) = Σ ΤλΓρφ^ί, g)] (1) FT is a Fourier transform: NT is usually 1. In Patent Reference 1, each mask pattern element is replaced by a point and a line, and a convolution calculus is performed with the interferogram to obtain an interferogram of the entire mask; therefore, the interferogram e (x, y) is presented. Simple coherence. However, the interferogram e ( x, y ) does not count the mask's pattern (such as the outline). When using an interferogram to calculate an approximate aerial image, the interferogram e'(x, y) in which the mask pattern is included must be obtained. To achieve this, a singular value © decomposition is performed on the TCC. Let λί be the ith eigenvalue * Φ s ( f, g ) be the ith eigenfunction, and a(f, g) be the diffracted light distributed over the mask pattern. Then use the following formula to find the interferogram e'(x, y) that is included in the mask pattern: Ν' _ e'(X, y) = HFT[a(f, g)CE>i(f, g) (2) i=l Use the interferogram e' ( x, y ) shown in relation (2) to generate an approximate aerial image. -17- 200931290 The following is an approximate aerial image calculation method that does not make a singular (intrinsic) decomposition of TCC. The mask pattern and wafer pattern in the semiconductor exposure device have a partially coherent image relationship, and the portion of the coherent image requires effective light source information to determine the homology of the mask surface. Coherence here means the degree of homology depending on the distance on the surface of the mask. The homology of the effective source is incorporated into the aforementioned TCC. The TCC is generally defined by the pupil plane of the projection optical system and represents the region where the effective source, the pupil function of the Φ optical system, and the complex conjugate of the pupil function of the projection optical system overlap each other. TCC is expressed as a four-dimensional function as follows: TCC(f', g,,f" , g")= JJs(f, g)P(f + f' , g + g- )p*(f + f" , g + g" )dfdg .-.(3) where (f, g) is the coordinate on the pupil plane, s(f, g) is a function describing the effective light source, and P(f, g) is the pupil function. Please note that * indicates a conjugate complex G number, and the integral range is from -〇〇 to 〇〇. The aberration of the projection optical system, the polarization of the illumination optical system, the photoresist information, and the like can be incorporated into the pupil function P(f, g). Within this specification, the term pupil function often contains the required polarization, aberrations, and photoresist. Use TCC to describe the function I(x,y) of the aerial image, and the four points of the TCC are as follows: y) = JJJJ TCC(f', g', f" , g" )a(f, g)a*(f ' , g') (4) x exp{- ί2π[(ί'-f" )x + (g'-g" )y]}df' dg' df"dg" -18- 200931290 where a(f,g To illustrate the Fourier transform function of the mask function, the function of the mask is a function of the spectral distribution (diffracted light distribution) on the mask. Please note that * indicates a conjugate complex number with a range of points from 00. For a detailed description of relation (4), please refer to M. Born and E. Wo If, "Principles of Optics", England, Cambridge University Press, 1 999, 7th (extend) edition, pages 554-632.

Assume that relation (4) is calculated directly by computer. In this case, using discontinuous variables, relation (4) can be rewritten as: y) = Σ a(f,' 5' ) βχρ[- ί2π(ί' x + g' y)] ... (5) f., g. x F'1[wf,gI(f,1, g")a*(f" , gH )] where F·1 represents the inverse Fourier transform. For fixed coordinates (f', g|), Wf,, g, (f", g") are defined as follows:

Wf,,g. = TCC(f,g',f",g") ...(6)

Since (f', g') is fixed here, Wf., g.(f", g") is a two-dimensional function and will be treated as a two-dimensional transmission crossover coefficient (TCC) within this specification. In computer calculations, the two-dimensional TCC is recalculated whenever an addition loop is performed and (f', g· ) is changed.

Wr,g’(f",g"). In relation (5), it is not necessary to describe the TCC as a four-dimensional function as in relation (3), and only two cycles are calculated. Therefore, 'by using the two-dimensional transmission cross coefficient' can shorten the calculation time and reduce the amount of calculation (to prevent the data capacity growth of the computer memory). The relation (5) can be rewritten as: -19- 200931290 Ι(Χ, Υ) = Yf, gl(x, y) ... (7) Μ· 1’,8’(17) is defined as:

Yf,,g,(x, y) = a(f·, g') exp[- ΐ2π(Γ x + g· y)] x F_1k.,g.(f",g")a,f", g")] ...(8 The aerial imagery method using relation (7) will be called Yf, g, (x, y) defined by the coordinates (f', g') of the aerial image decomposition method It is called a function to describe an element (airborne image element) of the aerial image in this specification. The difference between relation (3) and relation (6) will be explained in detail below. The center of the effective light source is assumed to be The origin of the pupil coordinate system. The TCC is defined as the sum of the movements of the pupil system P(f, g) coordinates (f', g') in the region, and the movement The function of the conjugate complex function p*(f,g) of the function p(f,g) and the function describing the effective source overlap with each other. p*(f,g) will often be A conjugate complex function called the pupil function P(f, g) of the projection optical system. Wr, g_(f", g") of relation (6) is shifted by a specified amount in p(f, g) ( f', g') is defined. \\^^(丨",§") is defined as having Light source and the pupil function overlap each other regions, and the effective light source and P * (f, g) is moved (f ", g ") obtained from the sum function of the overlapping area with one another.

Yf_,g'(x,y) is also defined when P(f,g) is shifted by a specified amount (f,,g|). |^, (factor 4) is multiplied by a conjugate complex function a*(f",g") that describes the spectral amplitude (diffractive light amplitude) on the mask, and the product is inverse FT--20- 200931290 Conversion. The function obtained by inverse Fourier transform, multiplied by a function exp[-i2n(f'x + g'y)] describing the oblique incidence effect corresponding to the displacement of the pupil function, and at coordinates ( f', g') diffracted light amplitude a(f', g'), yielding Yn(x, y). A conjugate complex function describing the amplitude of the diffracted light on the mask, which will be masked later The conjugate complex function of the diffracted light distribution is called. The function 0 exp[-i2n(f'x + g'y)] describing the oblique incidence effect will be explained below. Let 0 be the angle between the optical axis and the traveling direction of the plane wave. The plane wave traveling direction is described by exp[_i2n(f'x + g'y)]. Then, since sin2 0 = (NA/又)(f, 2, g'2), the plane wave traveling direction is inclined with respect to the optical axis, Therefore, the function describes the oblique incidence effect. exp[-i2n(rx + g'y)] can also be interpreted as a function to describe the plane wave traveling direction is a coordinate on the plane connecting the pupils. f|, g·) a line connecting the point where the optical axis intersects the image plane. Since the diffracted light amplitude a(r, g·) at the coordinates of the pupil plane (f', g, ) is constant, at coordinates The diffracted Q-light amplitude a(f',g] of (f,,g,) is multiplied by a multiplication-constant. The following is an approximation of the aerial image. The function describing the approximate aerial image, JappiXy, is defined as follows: M' Worker app(X' y) ® Σ Yf,,g,(X' y) · (9) ms*l * * * ' ' s円The coordinates (f ',g1 ) have a total of m combinations, M1 is equal to Or less than the integer of M, if, can get Μ · = Μ 'Approximate aerial image corresponds to the relationship (7) -21 - 200931290 perfect aerial image; if M, = l, approximate aerial image is Yf), Q ( x,y) , as the following relation: 1 Ι3ΡΡ(χ/ y) « Yf.,g.(X/ y) =, g")a*(f" , g" )] ...(10) —y〇,〇(x,y) where a(0,0) is a constant, which is the convolution integral between the function of the effective source and the conjugate complex function of the © pupil plane. Fourier transform and Anti-Fourier transforms are often exchanged and used, so 'describe there The convolution integral between the function of the light source and the pupil function or its conjugate complex function, multiplied by the diffracted light distribution or its conjugate complex function, and the resulting product is then subjected to Fourier transform or inverse Fourier transform to obtain Y〇,〇( x, y). In this way, 'using relation (9) or relation (10), assuming that the integer Μ 1 is equal to or smaller than M, approximate the aerial image to have one or two or several aerial image elements Yf', g, (X, y Add the resulting function definition. The following embodiment uses the aerial image elements Yf''g, (x, y) (relationship (ίο)) of M'=l in the hypothetical relation (9) to calculate an approximate aerial image; 2) Calculate approximate aerial imagery. The physical meaning of the approximate aerial image will be described in detail below. In coherent imaging, the image distribution function (a function describing the intensity distribution of the point image) can be determined. A positive position can be set as the open part in the point image distribution function, and a negative position can be set as the light shielding part (or the phase is 18). 〇. The open part) 'to make a Fresnei iens. If the Foucault lens manufactured by -22-200931290 is used as a mask, the contact hole can be isolated by coherent illumination. The Froh lens is based on the homologous illumination of the point image distribution function. The 'point image distribution function cannot be calculated in the partial coherent illumination. The amplitude on the image plane cannot be counted in the partial coherence imaging. The amplitude on the image plane is intrinsic. In the decomposition method, the physical meaning of the above-mentioned calculation method using the two-dimensional transmission cross coefficient is different. First, the point image distribution function should be characterized by the Fourier transform representation of the characteristic (modulation conversion function); the same frequency response characteristic is represented by the pupil function and the effective source (ie, the convolution integral between the pupil functions; the frequency response of the non-coherent illumination The optical 瞳 function is represented by autocorrelation. It is assumed to perform non-coherent illumination with homology σ = 1. The non-coherent illumination should be characterized by the effective light source of the pupil function. In view of this, the frequency response characteristics of partial coherent illumination are The convolution integral between 〇 and the effective light source is approximated; in other words, the frequency response characteristic is approximated by W0, Q(f", g". By simply converting WG, Q(f", and leaves, partial homology is obtained. The point image distribution function of the illuminating point image distribution function determines the opening shielding portion of the mask, and the contact hole can be exposed by the same effect as the Foucault lens. It is only necessary to calculate the point image distribution function and the mask function. Between the other, and based on the results obtained to determine the mask pattern, you can improve the arbitrary image effect. According to the relationship (10), the diffracted light distribution exposure shape Chengyizhi. However, this is because of the calculation. Therefore, the number of the 响 〇 and the intrinsic frequency of the frequency 照射 自身 自身 自身 自身 自身 自身 自身 自身 自身 自身 自身 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅 傅The product of the convolution integral mask and the product of -23- 200931290 w0,Q(f",g") is used as the Fourier transform to obtain Y〇,〇(x,y). The diffracted light distribution is a concealing function. Fourier transform ' We,Q(f",g") is the Fourier transform of the point image distribution function. Therefore ' is the convolution integral between the mask function and the point image distribution function according to the specified formula. It is equivalent to calculating the convolution integral of the point image distribution function and the mask function of the partial coherent imaging. As mentioned above, w〇, 〇(f", g") approximates the frequency 0 response characteristic of the coherent illumination. W f., g, (f", g" other than WG, G(f", g") may be an unformed frequency response characteristic that is omitted in the partial coherent illumination, similar to the frequency response characteristic. In addition to Y〇, 〇(x,y) other than ^4, (17) is in the calculation The component that is omitted when the point image distribution function and the mask pattern function of the partial coherent illumination are convolved. Therefore, when M' in the relation (9) is set to be larger than 1, the approximate accuracy is improved. The interferogram can be obtained by analyzing the intrinsic 四 of the four-dimensional TCC, so the aerial image calculation needs to square the absolute eigenfunction of the eigenfunction and add the result to Q. However, as shown in relation (7), the aerial image analysis method is used. To calculate aerial images, you only need to add the aerial image elements without the absolute square of the aerial image elements. The aerial image analysis method and the singularity (intrinsic 解析) analytical method use different physical quantities of different units, and thus have completely different characteristics. In step S104, the two-dimensional image is extracted from the near-air image calculated in step S102; precisely, by setting a reference slice (I.), a two-dimensional image of the approximate aerial image profile s e c t i ο η is extracted. For example, if the pattern data is a light transmission pattern, from the approximate aerial image, the extracted intensity 値 is equal to or higher than the predetermined 値 (which may be an arbitrarily set threshold ,), -24- 200931290 as the two-dimensional image. If the pattern data is in the image in the light shield, the portion whose intensity 値 is equal to or lower than the predetermined threshold , is taken as the two-dimensional image. In step S106, the control unit 20 compares the two-dimensional image of the step and the target pattern to determine whether the difference of the two-dimensional image falls within the preset tolerance. The parameter used to compare the pattern (measurement 値) can be the line width or the graph ^ 0 normalized image log slope, Normalized Image intensity peak 値. If the difference between the target patterns of the control unit 20 in step S1 06 falls within the preset allowable range, the main pattern that determines the difference between the two-dimensional image and the target pattern is determined. More precisely, in step S108, the control unit deforms the shape of the main pattern to determine a new main picture. The first to fourth embodiments will explain the detailed process at step S108. In step S1 1 0, using the pattern data made in step S108, the control unit 20 transmits information from the user, NA information, lambda information, aberration information, and light-like aerial image, and then returns the program. Go to step S1 04. Repeat the program of S 1 1 〇 until the 2D image and the target enter the preset allowable range. If the control unit 20 determines in step S106 that the difference between the two types falls within the preset tolerance, it is in the step pattern, and is self-approximate (may be a two-dimensional extraction between the target pattern and the target pattern S104). The image and target group length, NILS (Log Slop), or the determination of the two-dimensional image and the preset allowable 値 in step S108: 20 according to the two-dimensional image. The main figure that determines the detailed deformation of the main pattern will be described later. The effective light source resistance information of the type is calculated: the difference between the step S 1 0 4 and the pattern is determined by the falling image and the target figure S 112 is determined by the -25- 200931290 auxiliary pattern. First, by the step S110 In the approximate space calculated (that is, the approximate aerial image of the deformed main pattern), the position where the pattern is inserted (the auxiliary pattern insertion position) is extracted. The auxiliary pattern is set as a peak portion, and the peak portion is at the light intensity. The peak (local or local minimum 値) is neither in the area of the reference slice (I.) nor the overlap type (ie, in the area outside the area projected by the target pattern, please note that the true auxiliary pattern insertion position is Part of the mask and peak Q Then, the auxiliary pattern is determined according to the peak portion of the light intensity, and the auxiliary pattern is inserted into the peak portion. In the first embodiment to be described later, the detailed step of determining the auxiliary pattern in step S1 1 2 will be explained. In S114, the control unit 20 generates the main pattern selected in the step and the data of the auxiliary pattern selected in step S112 (the same auxiliary pattern is inserted as the mask data. In this manner, the mask data The generation program (mask data) performs the aerial image calculation step, the two-dimensional image extraction step, the φ determination step, the peak portion extraction step, the auxiliary pattern determination step, and the step. Please note that the auxiliary pattern insertion is often performed by the main pattern. The effect is changed. When this phenomenon occurs, the light approximation effect between the auxiliary pattern and the auxiliary pattern is only taken into account when the mask data is generated, as shown in FIG. 3 with reference to FIG. 3', in step S116, the control unit 20 sets the S108. The main pattern determined in the middle and the pattern obtained by the auxiliary addition determined in step S112 are used as pattern data to calculate the near image. In addition to the pattern information, the input information set is In the image in use, the auxiliary insertion bit is maximized in the main image field. The corresponding size is equal to the fourth!. Poly S 1 08 Incoming) Generates the main pattern and produces the light approximation main pattern No. In the step pattern Similar airborne input -26- 200931290 effect light source information, να information, lambda information, aberration information, and photoresist information. In step S118, control unit 20 extracts a two-dimensional image from the approximate aerial image calculated in step S116. In step S120, the control unit 20 compares the two-dimensional image extracted with the target pattern in step S118 to determine whether the difference between the two-dimensional image and the target pattern falls within the preset tolerance. Q If the control unit 20 determines in step S120 that the difference between the two-dimensional image and the target pattern falls outside the allowable threshold, the process returns to step s1. Then, the control unit 20 determines the main pattern that will cause the difference between the two-dimensional image and the target pattern to fall within the preset tolerance. If the control unit 20 determines in step S120 that the difference between the two-dimensional image and the target pattern falls within the allowable frame, the process proceeds to step S114. In step S114, the control unit 20 adds the main pattern determined in step S108 and the auxiliary pattern type which is the same as the one determined in step S112, to generate a pattern as the pattern data. In the flow shown in FIG. 3, the first aerial image calculation step corresponding to step S102 is first performed, the first two-dimensional image extraction step corresponding to step S104, and the first main pattern determination step corresponding to step S108. . Next, the first peak portion extracting step and the first auxiliary pattern determining step corresponding to step S112 are performed, corresponding to the second aerial image calculating step of step S116, corresponding to the second two-dimensional image extracting step of step S118. Finally, a second main pattern determining step corresponding to step S1 0 8 is performed, corresponding to the second peak portion extracting step and the second auxiliary pattern determining step -27-200931290 of step S112, and corresponding to step s 1 1 4 The steps of production. By using the mask generated by the procedure shown in Figures 2 and 3 for exposure processing and outputting the generated data, a high-precision micro-pattern can be formed on the substrate. In other words, the cover data has the possibility of forming a high-precision micro-pattern. The cover pattern can include a pattern produced by the above-mentioned mask data generation program. In the following first to fourth embodiments, the data generating program will be described in detail as a program for masking data, and the mask data to be interpreted. Note that λ is the wavelength of the exposed light, the number of apertures on the side of the image plane of the optical system, and the number of the object plane side of the number of aperture systems that the σ system directs to the surface of the mask. Since the pupil can set various flaws due to the wavelength λ of the exposed light and the number of apertures of the exposure apparatus, it is preferable to normalize the mask pattern ΝΑ /ΝΑ. For example, if the pattern of = 248 nm, ΝΑ = < 10 Onm is normalized by the aforementioned method, it is 0.29 state called k 1 conversion. The size of the pattern on the surface of the mask varies depending on the size of the pattern on the surface of the projection optics. To make it easy to simply zoom in on the projection optics in the example, multiply the mask size to set the pattern size on the surface of the mask to 1:1 between the wafer surface. By this setting, the comparison between the mask table and the coordinate system on the wafer surface is also 1: mask data, and the EB plotting device is produced to produce the mask. In addition to the generated mask, the NA generated by the program is the ratio of the self-illuminated light to the projected optical path. The projection optical system has a size of (λ).7 3 and a size of . The coordinate system of the first embodiment of the first embodiment is an enlarged view of the first embodiment of the first embodiment. The NA of the projection optical system used is 0.73 (corresponding to NA information) and the wavelength of the exposed light is 248 nm (corresponding to; I information). In addition, the projection optical system assumes that there is no aberration (corresponding to aberration information), and does not consider applying a photoresist (corresponding to photoresist information) on the wafer. The illumination light is not polarized. 0 The target pattern (pattern data) is the isolated line pattern as shown in Figure 4, with a line width of 120 nm and a line length of 2400 nm. In Fig. 4, the isolated line pattern is a light-shielding pattern (i.e., the transmittance is zero), and the transmittance in the region where the isolated line pattern does not exist (background) has a transmittance of 1. The phase is zero throughout the entire area. The graph of Fig. 4 shows the target pattern (pattern data) according to the first embodiment. The effective source uses quadrupole illumination as shown in Figure 5 (corresponding to effective source information). In Fig. 5, the white circle line indicates σ = 1, and the four white Φ areas indicate the light irradiation portion. Note that the graph of Fig. 5 shows an effective light source according to the first embodiment. Fig. 5 is only an example of an effective light source, and the present invention is not particularly limited thereto. The approximate aerial image shown in Fig. 6 is calculated from the target pattern and the aforementioned input information set (effective light source information, ΝΑ information, λ information, aberration information, and photoresist information) using the relation (10). In Figure 6Α, the target pattern is represented by a solid line that overlaps the approximate aerial image. In the approximate aerial image shown in Fig. 6Α, the isolated line pattern appears to have a circular arc edge, and the light shielding portion of the target pattern is short compared to the target pattern -29-200931290. The approximate aerial image shown in Fig. 6B is calculated by performing the aforementioned process 40 1, deforming the main pattern. In the aerial image of Figure 6B, the intensity distribution inside the target pattern is indicated by the solid line and is nearly uniform. Fig. 7A shows that the edge reference sheet (I.) 〇 95 I. 】 Cut the two-dimensional image obtained from the approximate aerial image of Figure 6 (0. Similarly, Figure 7B shows the edge reference film (U) 1.05 I. The two-sided image obtained by cutting the approximate aerial image of Figure 6B) . In Figures 7A and 7B, the target pattern is indicated by a solid line on the dimensional image. The reference slice (I.) is limited to the latter. The difference from the target pattern (for example, shape change, tilt 强度 or intensity peak 値, and log slope) can be calculated from the 2D image of Figure 7A and the approximate aerial image. The main pattern can be determined (deformed). The decision (deformation) of the main pattern will be explained in detail below. 8A shows 'cutting the target pattern and the pattern data (the same target pattern at the beginning), and cutting the target pattern and the cutting elements of the 2D image with the same number of cuts, and the difference deforms the pattern data ( Correction). Please pay attention to the target map. The pattern data can be eliminated by eliminating unnecessary elements or adding new elements. When the cutting elements of the pattern data are deformed according to the target pattern, the approximate data shown in the pattern data mask shown in Fig. 8B is obtained. Stacked on top of it 1.05 I. Cut section image) 0.95 1〇 and dimension image (cross-section overlapped on the two sides, also known as the degree of tilt, strong 7B according to the difference first, as shown in the figure type data and two-dimensional image. According to the two types Cut without deformation. Between 2D images. Then use the -30-200931290 deformed pattern data as the new pattern data to calculate the approximate aerial image. Repeat the same procedure until the target pattern and the 2D image The difference falls within the allowable enthalpy. As mentioned earlier, the approximate aerial image shown in Fig. 6B is calculated from the pattern data shown in Fig. 8B. Next, the auxiliary pattern is inserted. Approximate aerial from Fig. 6B Extracted from the image, in the region where the target pattern is not overlapped (that is, in the region outside the region projected by the main pattern), the light intensity is the peak portion of the peak. If the target Q pattern is a light-shielding pattern , calculate the peak portion in the region where the light intensity is equal to or higher than the threshold 且 and is darker than the background, and insert the square auxiliary pattern into the peak portion. When using the line pattern, only one dimension is required. Detection can be the peak In this case, the peak portion can be extracted only by detecting at least in a direction perpendicular to the longitudinal direction of the line pattern and in a direction parallel to the longitudinal direction of the line pattern. Inserting the auxiliary pattern The position may be the position of the center of gravity of the peak portion in a region darker than the background. However, if a line pattern of a certain size is used, the intensity of the main pattern affects the main pattern away from the approximate aerial image. Then, even in the area where the main pattern is not covered, the peak portion is covered with an intensity spike due to interference, so that the detection of the peak portion becomes difficult. In this case, the method of calculating the peak position (partial) It is also possible to calculate the second derivative 値 (positive or negative 値) of the light intensity distribution of the approximate aerial image (such as the quadratic differential sum (Laplace operator) along the two orthogonal axes). Figure 38 shows The second derivative (Laplacian) map of the intensity distribution. A differential (gradient 'gradient) along a particular direction has noise and is difficult to process. Therefore, a complex shape and a certain size of the line graph Type 'strong' The second derivative of the distribution is calculated to calculate -31 - 200931290. The peak portion is more effective. The line width dimension on one side of the auxiliary pattern must be small to be inaccurate. When using the light shielding pattern, the size background of one side The ratio of the light intensity to the minimum light intensity of the main pattern and the difference between the light intensity of the background and the area that does not overlap the target pattern. When using the line pattern, the auxiliary size is about its width. The 1/3 to u order of the line pattern size and the size of the main pattern are usually larger than the 値 when using the light transmission pattern. In the first embodiment, the size of the auxiliary pattern side is 40 nm. When the auxiliary pattern is inserted as described above, the pattern data of the deformed main image as shown in Fig. 9 is generated. Thus, the mask data shown in Fig. 9 is generated. It is confirmed (evaluated) whether the mask material is on the wafer surface. Form the desired aerial image. Mask data evaluation does not calculate the approximate aerial image, and calculates the aerial image. Figure 2 shows the second image of the aerial image. The pattern data (mask data) shown in Figure 9 is accurately calculated 10, the two-dimensional image is uniform, and its length is not reduced. For comparison, the calculation of the target pattern itself, and the aerial image of the image pattern itself inserted into the target pattern itself is shown in Figure 11 A itself, and the wide g pattern (scattered strip) is inserted into the target. The pattern obtained at halfway between the patterns (12 0nxn) is shown in Fig. 12A. Figure 1 is broken down. The fine calculation system is based on the difference between the two and the 1/2 of the light intensity minimum pattern, according to the exposure process, and the size of one side is in the width direction, and the pattern data is obtained. The mask is derived from accurate imagery. The reference picture is short, close to the main auxiliary pattern (.. for the target pattern f 40nm of the pitch (1B shows the 2D-32-200931290 image 'from the target pattern itself shown in Figure 11A) Aerial image. The two-dimensional image shown in Fig. 12B is an aerial image of the pattern obtained by inserting the scattered strip into the target pattern as shown in Fig. 12A. The pattern shown in Figs. 8B, 9, 11A' and 128 The two-dimensional image of the aerial image obtained by the data is 'quantitatively evaluated. More precisely, the aerial image is calculated by changing the out-of-focus image of the pattern data'. Therefore, in this embodiment, it is assumed that each figure is horizontal. The axis is the X-axis and the vertical axis is the y-axis. The position of the calculated line width is in the center of the 2D image (x = 〇, y = 0), and the center of the distance is 70% at y=1200 (x = 〇, y =84〇), and the distance from the center is 90% of y=1200 (x=〇, y=1080). Figure 13 shows the line width 'by the two-dimensional image of the aerial image according to the pattern data shown in Figure 8B. Calculated. The line width shown in Figure 14 is calculated from the two-dimensional image of the aerial image according to the pattern data shown in Figure 9. The line width of the display 15 is calculated from the two-dimensional image of the aerial image according to the pattern data shown in Fig. 11A. The line width shown in Fig. 16 is a two-dimensional image of the aerial image according to the pattern data shown in Fig. 12A. In Figures 13 to 16, the horizontal axis represents out-of-focus (Vm), and the vertical axis represents line width CD (nm). Note that the unit of size is nm. Referring to Figure 15, in accordance with Figure 1A In the two-dimensional image of the aerial image of the target pattern, the line width from the center of 90% is small, and the line width relative to the focus changes greatly. Referring to Fig. 16, the scattered strip is inserted into the target according to Fig. 12A. In the two-dimensional image of the aerial image of the pattern obtained by the pattern, the line width of the relative focus is relatively mild, but the line width from the center 70% and 90% is -33-200931290. Referring to Figure 13, According to the dimensional image of the deformed main aerial image shown in Fig. 8B, 70% of the distance from the center and the line width change little. Referring to Fig. 14', the image obtained by the auxiliary pattern main pattern data according to Fig. 9 is obtained. The two-dimensional image of the aerial image is separated from the center by 70% and 90% of the line width The minimum is φ, and the line width change is milder. In this mode, the image performance of the mask data generation program 401 mask data (pattern data) is better than the mask obtained by adding the dispersion strip generated by the operation. Therefore, the accuracy of the isolation line pattern. The size of the auxiliary pattern can be changed according to the light intensity of the peak part, instead of keeping it unchanged. More accurate shadow on the side of the i-aux auxiliary pattern The length, and a. is shown by reference to the length of the reference A, which changes the length ai of one side of the i-th auxiliary pattern. 17. Note that FIG. 17 is a schematic diagram explaining the change of the auxiliary pattern method. The first method is based on the ratio of the light intensity of the peak portion, the length ai of one side of the auxiliary pattern, as shown in the following equation: a± = a0 X 7ΪΓ : for the light transmission pattern Si = a〇X - Ιχ : The type insertion of the 90% position of the light-shielding pattern data is transformed, and the mask generated by the relative focus can be high-precision (the peak is ί, the ai is the size, There may be a method of changing the i-th-34- ... (11) 200931290 in the size of the figure, where Ii is the light intensity 値 at the position of the i-th auxiliary pattern, and Iback is the transmittance of the background. The pattern data of the 1 8 A display is the size of the auxiliary pattern in the pattern data (mask data) shown in Fig. 9. The second method is changed according to the inverse ratio of the light intensity of the peak portion. The length ai of one of the i auxiliary patterns is as shown in the following equation: ai = X > / ϊ7ϊ7 : for the transparent circle ❹ __ ai = X λ/1 / (Iback - I): Light Shielding Pattern...(12) In the second method, first determine the upper limit length, and then set ai in ai > a-limit. Figure 18 The pattern data displayed by B is the size of the auxiliary pattern in the pattern data (mask data) shown in Fig. 9. The second method is changed. 〇 In the first method, the approximate aerial image after inserting the auxiliary pattern The change before the distribution and insertion of the auxiliary pattern is not significant. Therefore, the change of the insertion-assisted pattern is not significant for the insertion-independent pattern, and the change in image performance (such as the increase in depth of focus) is small. In the method, the change of the approximate aerial image distribution after inserting the auxiliary pattern and the insertion of the auxiliary pattern are significant. Therefore, the second method helps to change the approximate aerial image distribution, so it is suitable for any operation to approximate the aerial image. For example The second method inserts a larger auxiliary-35-200931290 pattern at a low coherence position, which can increase the influence on the main pattern. In other words, it can be enhanced by inserting a larger auxiliary pattern at a position with low coherence. Coherence. It is also possible to enhance changes in image performance, such as increasing the depth of focus. Even if a larger auxiliary pattern is inserted at a low coherence position, the decision is lower than The upper limit of the auxiliary pattern that cannot be decomposed will never be improperly decomposed. In fact, the lower limit of the auxiliary pattern that can never be decomposed is determined by the approximate aerial image intensity of the auxiliary pattern insertion position. Therefore, the size of each auxiliary pattern can be calculated according to the intensity of the insertion position. Please note that the shape of the pattern does change significantly compared to when the auxiliary pattern is not inserted. If the shape of the pattern or its kind is overcorrected, Correction, the third embodiment will be explained. Second Embodiment The second embodiment assumes that the NA of the projection optical system used by the exposure apparatus is 0.75 (corresponding to NA information), and the wavelength of the exposed light is Q 193 nm ( Corresponds to information). In addition, the projection optical system assumes no aberration (corresponding to aberration information) and does not consider the photoresist applied to the wafer (corresponding to photoresist information). The illumination light is not polarized. The target pattern (pattern data) is a contact hole pattern as shown in Fig. 19, having a width of 120 nm, and arranged in an array at a half pitch of 100 nm (kl conversion 値 = 0.39). In Fig. 19, the contact hole pattern is a light transmission pattern (transmittance is 1), and the light transmittance of the region (background) having no contact hole pattern is zero. The phase is zero throughout the entire area. The graph of Fig. 19 shows the target pattern (pattern data) according to the second embodiment. -36- 200931290 The effective light source (corresponding to the effective light source information) using quadrupole illumination is shown in Figure 20. In Fig. 20, a white circle line indicates σ = 1, and four white areas indicate a light irradiation portion. Please note that the graph of Figure 20 shows an effective source of light in accordance with embodiment two. The approximate aerial image shown in Fig. 21 is calculated from the target pattern and the aforementioned information set (effective light source information, ΝΑ information, λ information, aberration information, and photoresist information) by the relation (10). In Figure 21, the target pattern is indicated by a solid line that overlaps the approximate aerial image. In the approximate aerial image shown in Fig. 2 1 , the light intensity peaks of the contact hole patterns are different. The approximate aerial image shown in Fig. 2 1B is obtained by executing the aforementioned mask data generating program 40 1, and deforming the main pattern to calculate. In the approximate aerial image shown in Figure 2 1 B, the light intensity peaks of the contact hole patterns are nearly identical. The decision (deformation) of the main pattern will be explained in detail below. Although the shape of the contact hole pattern is not complicated, only the size and position of the main pattern are changed without being cut, and the main pattern can be determined (deformed) by cutting the target pattern as in the embodiment. The deformation (correction) of the pattern data is based on the difference between the target pattern and the two-dimensional image of the approximate aerial image. Please note that the target pattern is not deformed. Then, the deformed pattern data is taken as the new pattern data, and the approximate aerial image is calculated. The same procedure is repeated again and again until the difference between the target pattern and the two-dimensional image falls within the tolerance. The approximate aerial image shown in Figure 2 1 B can be calculated from the pattern data thus obtained. Further, from the one-dimensional image of the approximate aerial image shown in FIG. 21B, -37-200931290' extracts a peak whose peak intensity is in a region where the target pattern is not overlapped (ie, a region other than the main pattern projection region). section. If each target pattern is a transmissive pattern, a peak portion of a region whose light intensity is equal to or lower than the threshold and brighter than the background is calculated, and a square auxiliary pattern is inserted into the peak portion. When using the contact hole pattern, the peak portion is extracted only by two-dimensional detection. The position at which the auxiliary pattern is inserted may be the position of the center of gravity of the peak portion in an area brighter than the background. Even when the contact hole pattern is used, the peak position (partial) can be easily calculated by counting the second derivative of the light intensity distribution (for example, the Laplacian) of the approximate aerial image. Figure 39 shows a second derivative (Laplacian) map of the intensity distribution. Moreover, even when a contact hole pattern of a certain size is used, the intensity of the main pattern has an effect on the position away from the main pattern in the approximate aerial image. Then, the peak portion is superimposed on the intensity peak caused by the interference even in the region where the main pattern is not overlapped, so that the detection of the peak portion becomes difficult. The line width of the dimension on one side of the auxiliary pattern needs to be small so as not to be decomposed. φ More precisely, when using the transmission pattern, the calculation of the size of one side is based on the ratio of the maximum 値 of the light intensity of the main pattern to the maximum 値 of the light intensity of the area of the non-overlapping target pattern. It. When using the contact hole pattern, the size of one side of the auxiliary pattern is about 60% to 80% of the line width of the contact pattern, depending on the exposure procedure and the size of the main pattern. In the second embodiment, the size of one side of the auxiliary pattern is 75 nm. When the aforementioned auxiliary pattern is inserted into the deformed main pattern, the pattern data as shown in Fig. 22 can be obtained. The pattern data shown in Fig. 22 is thus generated as a mask material. Confirm (evaluate) the mask made from the mask material is -38-

200931290 No The desired aerial image can be formed on the surface of the wafer. The two-dimensional image of the approximate aerial image shown in Fig. 23 is accurately calculated from the pattern data (mask data) of Fig. 22. The two-dimensional image of the aerial image shown in Fig. 24B is calculated from the target pattern itself shown in Fig. 24A. Comparing Fig. 23 with Fig. 24B, it is found that the two images shown in Fig. 23 are more uniform and have less elliptical distortion than the two-dimensional image shown in Fig. 24B. A two-dimensional image of the null image calculated from the pattern data shown in Figs. 22 and 14A is quantified. More precisely, the aerial image is calculated by changing the out-of-focus of the pattern, so the line width (hole diameter) is calculated. In the right embodiment, the contact hole pattern CH2 at the edge of the hole array of the contact hole distribution in the center of the two-dimensional image is isolated. The line width CD of the contact hole pattern CH3 in the middle of the I column is as shown in Fig. 24A. The line width shown in Fig. 25 is calculated from a two-dimensional image of the aerial image according to the pattern shown in Fig. 22. The line width shown in Fig. 26 is a two-dimensional image of an aerial image calculated from the pattern data shown in Fig. 24A. In the figure and 26, the 'abscissa indicates the out-of-focus (^m), and the ordinate indicates the line width. According to FIG. 26, the contact hole pattern is isolated in the two-dimensional image of the image according to the target pattern itself shown in FIG. 24A. The line width of the contact pattern of the type and the array of holes in the contact cloth varies greatly, and the line width of the relative focus changes greatly. Referring to FIG. 25, in the same manner as shown in FIG. 22, the two-dimensional shadow of the aerial image of the pattern obtained by inserting the auxiliary pattern into the main pattern of the deformation is shown as: 1: According to the 25 CD air distribution, the line width variation is small, and its line width change relative to the focus is also small. In this manner, the image performance of the mask data (pattern data) generated by the concealer data generating program 4〇1 is superior to the image performance of the mask data generated by the learned technology. This results in a more accurate contact hole pattern. As with the first embodiment, the size of the auxiliary pattern can be changed depending on the light intensity of the peak portion (the size of the peak ,), rather than being maintained. The pattern data shown in Fig. 27 is obtained by changing the size of the auxiliary pattern in the pattern data (mask data) shown in Fig. 22 by the second method. By varying the size of the auxiliary pattern, it is possible to improve the effect of the auxiliary pattern on the contact pattern of the isolated contact hole or the contact hole pattern at the edge of the array of contact holes. The line width shown in Fig. 28 is calculated from the isolated contact hole pattern in the center of the two-dimensional image of the aerial image according to the pattern data shown in Figs. 22 and 27. As described above, the pattern data shown in Fig. 22 is obtained by maintaining the auxiliary pattern type + size fixed and the figure shown in Fig. 27 is changing the size of the auxiliary pattern. In Fig. 28, the abscissa indicates out-of-focus (the ordinate indicates the line width CD (nm). Referring to Fig. 28, when the size of the auxiliary pattern is changed (Fig. 27), the change in line width with respect to focus is smaller (i.e., the focusing characteristics are smaller). Good) When the size of the auxiliary pattern remains fixed (Fig. 22). The third embodiment is as described above for the insertion of the auxiliary pattern, and the new approximation effect between the auxiliary pattern and the main-40-200931290 pattern Especially when the size of the auxiliary pattern changes according to the inverse ratio of the light intensity of the peak portion, the change in the homology is significant. Therefore, it is possible to accurately change the shape of the pattern and the shape of each pattern. In this case, as shown in the figure. As shown in the flow chart of 3, after inserting the auxiliary pattern into the deformed main pattern, it is effective to calculate the approximate aerial image again, and further deform (correct) the pattern data. The example of Example 1 illustrates the third embodiment. The approximate aerial image shown in Figure 29A is calculated prior to inserting the auxiliary pattern into the deformed main pattern of the first embodiment. Figure 29B shows the approximate aerial image system. The calculation is performed after inserting the fixed size auxiliary pattern into the deformed main pattern of the first embodiment. Comparing the approximate aerial images of Figs. 29A and 29B, the display shows a significant change in the distribution. The approximate aerial image shown in Fig. 29B Extracting the two-dimensional image, and transforming the main pattern according to the two-dimensional image, as in the first embodiment, and then inserting the auxiliary pattern into the main pattern. If no new auxiliary pattern is extracted, it is available, or The same auxiliary pattern is not used. In this case, since the shape of the main pattern is changed, the light approximation effect changed by the auxiliary pattern insertion can be corrected. Then the deformed pattern data is used as the new pattern. Data, calculate the approximate aerial image. Repeat the same procedure until the difference between the target image and the 2D image falls within the allowable frame. Figure 3 shows the pattern data (mask data) thus obtained. The two-dimensional image of the aerial image calculated from the pattern data shown in Fig. 30 is used for quantitative evaluation. More precisely, the aerial image is calculated by changing the out-of-focus data of the pattern data. In the embodiment, it is assumed that the horizontal axis is the X-axis and the vertical axis is the y-axis, and the position of the line width is calculated in the center of the two-dimensional -41 - 200931290 image (x = 〇, y = 0), and the distance is in the center. 70% of y=1200 (x = 〇, y = 840), and the center of the distance is 90% of y=1 200 (x = 0, y = 1 0 8 0 ). Figure 31 shows the line width 'by Calculated according to the two-dimensional image of the aerial image of the graphic data shown in Fig. 30. Comparing Figs. 15 and 31 shows 'in the two-dimensional image of the aerial image according to the graphic data shown in Fig. 30' The change is small, and the line width change of the relative focus is warmer than that shown in Fig. 11. The line width calculation result shown in Fig. 31 is almost the same as that shown in Fig. 14. Shortening the length is a serious problem when using line patterns. To solve this problem, the edge of the line graph must be evaluated. In the best focus state, the line pattern is only extended in accordance with the expected line pattern shortening. However, the line pattern is shortened with the out of focus, so the length is not changed by focusing. Also, the comparison of the edges of the online graphics should be good. The comparison of the edges of the online graphs is based on the NILS assessment. Figure 32 shows the result of dividing the line pattern length by degrees before focusing in the best focus state, and checking for changes due to focusing. Figure 33 shows the results of this NILS calculation at the edge of the line graph and checks for changes due to focus. In FIG. 32 and FIG. 33, Pattern_before indicates that the pattern data is the target pattern itself (ie, the initial pattern data, FIG. 11A), and the pattern data indicated by the SB is the pattern obtained by inserting the scattered strip into the target pattern. (Fig. 12A) = The pattern data indicated by the PC indicates the pattern obtained by deforming the main pattern according to the approximate aerial image (Fig. 8B). The pattern data of the °OPCl+assistl word is based on the first The embodiment 'deforms the main pattern according to the approximate aerial image, and inserts the auxiliary pattern with a fixed size-42-200931290. The main pattern (Fig. 9) = OPC2 + aSSist2 indicates the status of the pattern is based on the In a third embodiment, the main pattern is deformed according to the approximate aerial image and an auxiliary pattern having a fixed size is inserted into the main pattern (Fig. 30). Referring to Fig. 32, in the respective conditions of OPCl+assistl (Fig. 9) and 〇PC2 + assist2 (Fig. 30), the line pattern length change due to the change in focus is minimized. Referring to Figure 33, in the best focus state of OPC2 + aSsist2 (Figure 30), the nils at the edge of the line pattern is the largest. Moreover, the focus of the NILS of the line graph edge relative to OPCl+assistl (Fig. 9) is almost unchanged, but the condition of PC2 + assist2 is weak. Fourth Embodiment An embodiment in which a line pattern has a target pattern of another shape will be explained below. The fourth embodiment assumes that the NA of the projection optical system used by the exposure apparatus is 0.73 (corresponding to NA information), and the wavelength of the exposed light is 〇 1 93 ηιη (corresponding to λ information). In addition, the projection optical system uses the focus position shift (out of focus) as the aberration of the projection optical system, but does not consider the photoresist applied to the wafer (corresponding to the photoresist information). The illumination light is not polarized. The target pattern (pattern data) is shown in Figure 35, which is an L-shaped (elbow) pattern with a vertical and horizontal dimension of 12OOnm. In Fig. 35, the isolated line pattern is a light-shielding pattern (transmission is zero), and the area (background) where no isolation line pattern exists is light transmittance. The phase is zero in the entire region. The graph of Fig. 35 shows a target pattern (pattern data) according to the fourth embodiment. The effective source uses four-pole illumination (corresponding to the effective source -43- 200931290 information), as shown in Figure 5. The approximate aerial image is calculated from the target pattern and the aforementioned input information set (effective light source information, NA information, lambda information, aberration information, and photoresist information) by a relation (10). A two-dimensional image is extracted from the approximate aerial image calculated. The pattern data is calculated based on the difference between the target pattern and the two-dimensional image of the approximate aerial image. In the region where the target pattern is not overlapped, the peak portion of the peak is sequentially extracted from the image in the approximate space calculated from the pattern data, so the auxiliary pattern is inserted into the peak portion. Figure 36 shows the pattern data thus generated as a mask material. Figures 37Α and 37Β show aerial images that are accurately calculated from the mask data. Figure 37 shows that the aerial image is formed in the best focus state; the aerial image shown in Figure 37Β is formed at a defocus of 0.2/zm. Referring to Figures 37A and 3 7 B, the corners of the aerial image obtained are uniform and the distortion is small, so that the performance of the image projected onto the wafer is suppressed from deteriorating when the focus is out of focus. In other words, the depth of focus increases, resulting in improved image performance.曝光 The exposure apparatus 1 is explained below with reference to FIG. The schematic block diagram of Fig. 34 shows the configuration of the exposure apparatus. The mask 130 used is manufactured based on the mask data generated by executing the mask data generating program. The exposure apparatus 100 is an immersion exposure apparatus that exposes a pattern of the mask 130 to the wafer 150 via a liquid LW between the wafer 150 and the projection optical system 丨40. Although the exposure apparatus 100 employs the step and scan mechanism of this embodiment, the step and repeat mechanism, or other exposure mechanism, can also be employed. As shown in FIG. 34, the exposure apparatus 10 includes a light source 110, an illumination light-44-200931290 system 120', a mask holder 135 for mounting the mask 130, a projection optical system 140, and a wafer 1150. A wafer rack 150, a liquid supply/recovery unit 160, and a main control system 170. The light source 11A and the illumination optical system 12' constitute a mask 130 on which the illumination device's illumination pattern is to be transferred. The light source 110 is an excimer laser such as a KrF excimer laser having a wavelength of 248 nm or an ArF quasi-fraction Q sub-clock of a wavelength of I93 nm. However, the type and number of the light source 11 are not particularly limited. For example, an F2 laser with a wavelength of approximately 157 nm can also be used as the light source 1 1 0. The illumination optical system 120 emits the mask i 3 0 with light from the light source 1 1 。. Illumination optics 1 20 can provide various illumination modes, such as conventional illumination and varying illumination (e.g., quadrupole illumination). In this embodiment, the illumination optical system 120 includes a beam shaping optical system 121, a collecting optical system 122, a polarization control unit 123, an optical integrator 124, and an aperture stop 125. The illumination optical system 120 also includes a condenser lens 126, a bending mirror 127, a masking blade 128, and an imaging lens 129. The beam shaping optical system 121 may be, for example, a beam expander including a plurality of cylindrical lenses. The beam shaping optical system 1 2 1 converts the horizontal-to-vertical ratio 剖面 of the cross-sectional shape of the collimated light from the light source 110 into a predetermined chirp (for example, converting the cross-sectional shape from a rectangle to a square). In this embodiment, the beam shaping optics 1 1 1 aligns the light from the source 110 into the size and spread angle required to illuminate the optical integrator 1 24 -45 - 200931290 degrees. The concentrating optical system 122 includes a plurality of optical elements and effectively directs light shaped by the beam shaping optical system 121 to the optical integrator 124. For example, the concentrating optical system 122 includes a zoom lens system that adjusts the shape and angle of light entering the optical integrator 124. For example, the polarization control unit 1 2 3 includes a polarization element disposed at a position 共 conjugated to the pupil plane 142 of the projection optical system 140. The polarization control unit 123 controls a polarization state of a predetermined region of the effective light source formed on the pupil plane 142 of the projection optical system 140. The function of the optical integrator 124 is to illuminate the illumination mask 130. The sentence is averaged, the angular distribution of the incident light is converted into a positional distribution, and the obtained light is output. For example, the optical integrator 124 has a fly-eye lens that maintains a Fourier-converted relationship between the incident surface and the exit surface. The fly-eye lens is composed of a combination of a plurality of rod lenses (i.e., microlenses). However, the optical integrator 124 is not particularly limited to a fly-eye lens, and may be, for example, an optical rod, a diffraction grating, or a cylindrical lens array plate. The aperture limiter 125 is set at a position ' immediately after the exit surface of the optical integrator 124' and is conjugated to an effective light source formed on the pupil plane 142 of the projection optical system 140. The aperture shape of the aperture limiter 125 corresponds to the light intensity distribution (i.e., effective source shape) of the effective source formed on the pupil plane 142 of the projection optical system 140. In other words, the aperture limiter 1 2 5 controls the light intensity distribution of the effective light source. The aperture limiter 125 can be switched in accordance with the illumination mode. Regardless of whether or not the aperture limiter -46-200931290 is used, the effective light source shape can be adjusted by a diffractive optical element (CGH) or 设定 set on the light source side of the optical integrator 124. The condensing lens 126 shrinks the light emitted from the second light source formed near the exit surface of the optical integrator 124, passes through the aperture limiter 125, and uniformly illuminates the light through the refracting mirror 127 to the mask sheet 1 2 8 . The mask sheet 128 is set at a position conjugate with the mask 130, and is formed by a plurality of movable light shielding plates. The mask sheet 128 forms a nearly rectangular opening corresponding to the effective area of the projection optical system 140. Light passing through the mask sheet 128 is used as irradiation light to illuminate the mask 130. The imaging lens 129 forms a light image that passes through the opening of the mask sheet 128 on the mask 130. The mask 130 is manufactured according to the mask data generated by the processing apparatus 1 described above, and has a circuit pattern (main pattern) to be transferred and an auxiliary pattern. The cover 130 is supported and driven by the shield frame 135. The diffracted light generated by the mask 130 is projected onto the wafer 150 by the projection optical system 140. The mask 130 and the crystal 150 are set in an optically conjugate relationship. Since the exposure device 100 is the stepping and scanning mechanism, the circuit pattern of the mask 130 to be transferred is synchronously scanned to be transferred to the wafer 150» when the exposure device 1 is the stepping and repeating mechanism. When the mask 130 and the wafer 150 are stationary, exposure is performed. The mask frame 135 supports the mask 130 by a masked chuck and is connected to a driving mechanism (not shown). The drive mechanism (not shown) is exemplified by a linear motor formation and is driven in the X-, y-, z-axis, and in the direction of rotation about each of the axes. Note that the scanning direction of the mask 130 and the wafer surface 47 - 200931290 is defined as the y-axis direction, and the direction perpendicular thereto is defined as the X-axis direction, which is perpendicular to the surface of the mask 130 or the wafer 150. Defined as the z-axis direction. Projection optics 140 projects the circuit pattern of mask 130 onto wafer 150. The projection optical system 140 may be a dioptric system, a catadioptric system, or a catotric system. The most final projection lens (final surface) of the projection optical system 140 is coated with a coating to reduce the influence of the liquid LW supplied by the liquid supply/recovery unit (for protection). The wafer 150 is a substrate on which the circuit pattern of the mask 130 is projected (shifted). However, the wafer 150 can be replaced by a glass plate or other substrate. Wafer 150 is coated with a photoresist. The wafer holder 155 supports the wafer 150, and uses a linear motor to move the wafer 150 in the X-, y-, z-axis, and in the direction of rotation about each of the axes, the same as the mask © frame 1 3 5 . The function of the liquid supply/recovery unit 1 60 supplies the liquid LW to the space between the wafer 150 and the final lens (final surface) of the projection optical system 140. The liquid supply/recovery unit 1 has a function of recovering a liquid supplied to a space between the wafer 150 and the final lens (final surface) of the projection optical system 1 40. A material having a higher transmittance than the exposure apparatus, preventing dust from adhering to the projection optical system 140 (on the final lens), and matching the photoresist treatment is selected as the liquid LW. -48- 200931290 The main control system 170 includes a CPU and a memory, and controls the operation of the exposure device. For example, the main control system 170 is electrically coupled to the mask holder 135, the wafer holder 155, and the liquid supply/recovery unit 160, and controls the synchronous scanning between the mask holder 135 and the wafer holder 155. The main control system 170 controls the supply, recovery, and stop supply/recovery switching of the liquid LW based on the scanning direction and rate of the wafer holder 155 during exposure. The main control system 1 70 performs these controls, in particular based on the information input from the monitor and the input device, i.e., information from the illumination device. For example, the main control system 170 drives the aperture limiter drive by the drive mechanism. Control information and other information of the main control system 1 〇 are displayed on the monitor and the monitor of the input device. The primary control system 170 receives information about the effective light source and controls, such as the aperture limiter, diffractive optical element, and chirp, in accordance with one of the foregoing embodiments, thereby forming the effective light source. At the time of exposure, the light beam emitted from the light source 110 illuminates the mask 130 by the illumination optical system 120. The projection optical system 140 passes the light beam through the mask 130 via the liquid LW, reacts the circuit pattern thereon, and forms an image on the wafer 150. The exposure device 1000 has excellent imaging performance and can provide high productivity and good economic benefits for devices such as semiconductor devices, LCD devices, image sensing devices (such as CCD), and thin film magnetic heads. The manufacturing steps of these devices are a step of exposing 100 substrates (such as wafers or glass plates) coated with a photoresist (photosensitive agent) using the exposure device, and developing a substrate to be exposed, And other familiar steps. Although the present invention has been described with reference to the embodiments, it is to be understood that the invention is not to be construed as limited. Structure and function. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic block diagram showing the structure of a processing apparatus which performs a method for generating a method according to one aspect of the present invention. The flowchart of Fig. 2 illustrates a mask generating data program which executes a mask data generating program by the control unit of the processing device of Fig. 1. D. FIG. 3 is a flow chart explaining another process for generating a mask data. The other program executes a mask data generation program by the control unit of the processing device of FIG. 1. FIG. 4 is a diagram showing the target pattern according to the first embodiment. (Graphic data). The graph of Fig. 5 shows an effect source 0 according to the first embodiment. The graphs of Figs. 6A and 6B show near-simultaneous aerial images calculated in the first embodiment. 7K of Figures 7A and 7B are intended to show two-dimensional images extracted from the approximate aerial image of Figures 6A and 6B. The graphs of Figures 8A and 8B explain the deformation of the pattern data (main pattern). Figure 9 is a diagram showing the insertion of the auxiliary pattern into the main pattern of the first embodiment, and the obtained pattern data. Figure 10 is a schematic view showing a two-dimensional image of a precise aerial image obtained from the pattern data shown in Figure 9. . -50- 200931290 The graph of Fig. 11A shows the target pattern according to the first embodiment. Figure 11B is a schematic diagram showing a two-dimensional image of an aerial image calculated from the target pattern shown in Figure 11A. The graph of Fig. 12A shows the pattern obtained by inserting the dispersed strip into the target pattern according to the first embodiment. Figure 12B is a schematic diagram showing a two-dimensional image of an aerial image calculated from the pattern shown in Figure 12A. 0 The graph of Fig. 13 shows the calculated line width from the two-dimensional image of the aerial image according to the pattern data shown in Fig. 8B. The graph of Fig. 14 shows the calculated line width from the two-dimensional image of the aerial image according to the pattern data shown in Fig. 9. The graph of Fig. 15 shows the calculated line width from the two-dimensional image of the aerial image according to the pattern data shown in Fig. 11A. The graph of Fig. 16 shows the calculated line width from the two-dimensional image of the aerial image based on the pattern data shown in Fig. 12A. 〇 The schematic of Figure 17 illustrates a method of changing the size of an auxiliary pattern. The graphs of Figs. 18A and 18B show the pattern data obtained by changing the size of the auxiliary pattern in the pattern data shown in Fig. 9. The graph of Fig. 19 shows a target pattern (pattern data) according to the second embodiment. Figure 20 is a chart showing an effective light source in accordance with a second embodiment. The graphs of Figures 21A and 21B show approximate aerial images calculated in the second embodiment. Fig. 22 is a diagram showing the pattern data obtained by inserting the auxiliary pattern into the main type -51 - 200931290 of the second embodiment. Figure 23 is a schematic diagram showing a two-dimensional image of an accurate aerial image taken from the pattern data shown in Figure 22. The graph of Figure 24A shows a target pattern in accordance with a second embodiment. Figure 24B is a schematic diagram showing a two-dimensional image of an aerial image calculated from the target pattern shown in Figure 24A. The graph of Fig. 25 shows the calculated line width from the two-dimensional image of the 空中 aerial image based on the pattern data shown in Fig. 22. The graph of Fig. 26 shows the line width calculated from the "two-dimensional image of the aerial image according to the pattern data shown in Fig. 24A". The graph of Fig. 27 shows the pattern data obtained by changing the size of the auxiliary pattern in the pattern data shown in Fig. 22. The graph of Fig. 28 shows the isolated contact point pattern from the center of the two-dimensional image of the aerial image according to the pattern data shown in Figs. 22 and 27, and the calculated line width is calculated. Ο The graph of Fig. 29A shows the approximate aerial image calculated before the auxiliary pattern is inserted into the main deformation pattern in the first embodiment. The graph of Fig. 2B shows the approximate aerial image calculated after inserting the fixed size auxiliary pattern into the deformed main pattern in the first embodiment. The graph of Fig. 30 shows the pattern data obtained by inserting the auxiliary pattern into the main pattern of the third embodiment. The graph of Fig. 31 shows the calculated line width from the two-dimensional image of the aerial image according to the pattern data shown in Fig. 30. The graph of Fig. 32 shows the result of 'severing the length of the -52-200931290 line pattern' before focusing to the best focus state and checking for changes due to focusing. The graph of Figure 33 shows the NILS results at the edge of the line graph and checks for changes due to focus. Figure 34 is a schematic block diagram showing a device-oriented configuration in accordance with one aspect of the present invention. The graph of Fig. 35 shows the target pattern data according to the fourth embodiment). 〇 The schematic diagram of Figure 36 shows the pattern data obtained by inserting the auxiliary pattern into the fourth implementation pattern. The schematic of Figure 37A shows a precise aerial image of the mask pattern shown in Figure 36. The schematic of Figure 37B shows an aerial image at the time of defocusing. The graph of Figure 38 shows the approximate aerial image differentiation shown in Figure 6B. The graph of Figure 39 shows the approximate aerial image subdifferential shown in Figure 21B. [Main component symbol description] 1 : Processing device 1 〇: Bus line 20: Control unit 3 〇: Display unit 40 _· Storage unit calculates the exposure of the junction (the second image of the second image obtained by the eye of the pattern example) 53- 200931290 50 : Input unit 60 : Media interface 401 : Mask data generation program 4 0 8 : Mask data 4 1 0 : 2D image 70 : Storage medium 4 0 2 : Pattern data © 403 : 404 : 405 : 406 : 407 : 409 : 4 11: 100 : 〇 13 0: 150: 110: 120 : 13 5: 140 : 15 5: 160: 170: Effective light source information NA information λ information aberration information photoresist information approximate aerial image was Deformation pattern data exposure equipment mask wafer light source illumination optical system mask frame projection optical system wafer holder liquid supply / recovery unit main control system -54 200931290 121 : 122 : 123 : 124 : 125 : 126 : 127 : © 128 ·· 129 : 142 : Beam shaping optics concentrating optics polarization control unit optical integrator aperture limiter concentrating lens refracting mirror mask imaging lens pupil plane ❹ -55-

Claims (1)

200931290 X. Patent application scope 1. A method for generating mask data for use by an exposure device by using a computer, the exposure device comprising: an illumination optical system for illuminating the mask with light from a light source, and projecting a pattern of the mask a projection optical system on a substrate, the method comprising: an aerial image calculation step, according to the illumination optical system, forming information related to a light intensity distribution of the pupil on a pupil plane of the projection optical system, Information about the wavelength of the light from the light source, information about a numerical aperture of the projection optical system on one side of the image plane of the projection optical system, and a target pattern to be formed on the substrate, Calculating an aerial image formed on an image plane of the projection optical system; a two-dimensional image extraction step of extracting a two-dimensional image from the aerial image calculated in the aerial image calculation step; the main pattern determining step, according to the second The two-dimensional image extracted by the dimension image extraction step determines the main pattern of the mask; the peak portion Extracting step, extracting a peak portion from the aerial image calculated in the aerial image calculation step; the peak portion is in a peak region except for the region where the main pattern is projected on the image plane; the auxiliary pattern Determining a step of determining an auxiliary pattern according to the light intensity of the peak portion extracted in the peak portion extraction step; and generating a step of inserting the auxiliary pattern determined by the auxiliary pattern determining step into the mask a portion of the cover corresponding to the peak portion extracted during the peak portion extraction step, thereby generating a map including the auxiliary pattern and the main pattern determined within the main-56-200931290 pattern determining step Type data as the material for the mask. 2. The method of claim 1, wherein in the two-dimensional image extraction step: if the target pattern is a light transmission pattern, extracting an area of the aerial image whose light intensity is not less than a critical threshold; and if If the target pattern is a light-shielding pattern, the area where the light intensity is not greater than the critical threshold is extracted. 3. The method of claim 1, wherein in the main pattern determining step, the main pattern is deformed until a difference between the two-dimensional image and the target pattern falls within a permissible range . 4. The method of claim 3, wherein one of the light intensity, NILS, and line width of the aerial image is used as a measure of the difference between the two-dimensional image and the target pattern. 5. The method of claim 1, wherein in the auxiliary pattern determining step, if the target pattern is a light transmission pattern, the light intensity according to the main pattern is the largest and the main pattern The ratio of the maximum intensity 値 of the area projected outside the area of the substrate determines the size of the auxiliary pattern; and if the target pattern is a light shielding pattern, the light intensity according to the background and the main pattern The difference between the minimum intensity of the light intensity and the difference between the light intensity of the background and the minimum light intensity of the region outside the region where the main pattern is projected onto the substrate determines the auxiliary pattern size. 6) The method of claim 1, wherein in the extraction step -57-200931290, the second derivative is calculated for the aerial image calculated in the aerial image calculation step, and the second derivative is obtained according to the acquisition Extract the peak portion. 7. A method of using a computer to generate mask data for use in an exposure apparatus, the exposure apparatus comprising: an illumination optical system that illuminates the mask with light from the light source, and projection optics that project the pattern of the mask onto the substrate a system, the method comprising: a first aerial image calculation step, according to information related to a light intensity distribution formed by the illumination optical system on a pupil plane of the projection optical system, and information about wavelengths of light from the light source, An information about a number of apertures on one side of the image plane of the projection optical system, and a target pattern formed on the substrate, and calculating an aerial image formed on an image plane of the projection optical system; a dimension image extraction step of extracting a two-dimensional image from the aerial image calculated in the first aerial image calculation step; a first main pattern determining step, according to the first two-dimensional image extraction step Ο a dimensional image that determines a main pattern of the mask; a first peak portion extraction step from which the air is calculated in the first aerial image calculation step Extracting a peak portion; the peak portion is light peak in addition to the region in which the main pattern is projected on the image plane; the first auxiliary pattern determining step is extracted according to the first peak portion extracting step The light intensity of the peak portion determines an auxiliary pattern; the second aerial image calculation step is related to the light intensity distribution formed by the illumination optical system on the pupil plane of the projection optical system according to a pattern Information, the information related to the wavelength of the light from the light source - 58 - 200931290, and the information related to the number of apertures on the image plane side of the projection optical system, and calculating the image in the projection optical system An aerial image formed on a plane; the pattern includes the main pattern determined in the first main pattern determining step and the auxiliary pattern determined in the first auxiliary pattern determining step, the auxiliary pattern And inserting a portion of the mask corresponding to the peak portion extracted in the first peak portion extraction step; a second two-dimensional image extraction step, calculating a step from the second aerial image Extracting the two-dimensional image from the aerial image calculated in the step; the second main pattern determining step determines the main pattern of the mask according to the two-dimensional image extracted by the second two-dimensional image extracting step; a second peak portion extraction step of extracting a peak portion from the aerial image calculated in the second aerial image calculation step; the peak portion is in addition to the region where the main pattern is projected on the image plane, and the light intensity is a second auxiliary pattern determining step of determining an auxiliary pattern according to the light intensity of the peak portion extracted by the second peak portion extraction step; and a generating step, the second auxiliary pattern determining step Determining the auxiliary pattern into a portion of the mask, the portion corresponding to the peak portion extracted in the second peak portion extraction step; thereby generating the auxiliary pattern and the second main pattern Determining the pattern data of the main pattern determined in the step as the material of the mask. A mask manufacturing method comprising: manufacturing a mask according to the data generated by the method of producing any one of claims 1 to 7. 9. An exposure method comprising the steps of: -59-200931290 manufacturing a mask by the mask manufacturing method of claim 8; irradiating the manufactured mask; and masking the mask with the projection optical system A picture of the image is projected onto a substrate. 10. A device manufacturing method comprising the steps of: exposing a substrate using an exposure method of claim 9; and developing a substrate to be exposed ❹ 1 1. a storage medium, a program for storing Having the computer generate masking data for use by the exposure apparatus; the exposure apparatus includes: an illumination optical system that illuminates the mask with light from the light source, and a projection optical system that projects the pattern of the mask onto the substrate, the medium Having the computer perform the following steps: an aerial image calculation step, based on information related to a light intensity distribution formed on the pupil plane of the projection optical system, information related to the wavelength of light from the light source, at the projection The image of the optical system is related to a number of apertures on one side of the image plane, and a target pattern formed on the substrate, and an aerial image formed on the image plane of the projection optical system is calculated; Extracting step of extracting a two-dimensional image from the aerial image calculated in the aerial image calculation step; a pattern determining step of determining a main pattern of the mask according to the two-dimensional image extracted by the two-dimensional image extraction step; a peak portion extraction step of extracting a peak portion from the aerial image calculated in the aerial image calculation step The peak portion is in the peak region except for the main pattern projection -60-200931290 in the image plane; the auxiliary pattern determining step is based on the peak portion extracted in the peak portion extraction step. The light intensity determines an auxiliary pattern; and a generating step of inserting the auxiliary pattern determined by the auxiliary pattern determining step into a portion of the mask, the portion extracted from the peak portion extraction step The peak portion corresponds to generate the pattern data including the auxiliary pattern and the main pattern determined in the main pattern determining step as the mask material. 1 2 · a storage medium, the stored program causes the computer to generate mask data for use by an exposure device, the exposure device comprising: an illumination optical system that illuminates the mask with light from a light source, and a pattern of the mask a projection optical system projected on a substrate, the medium causing the computer to perform the following steps: a first aerial image calculation step, based on information related to a light intensity distribution formed on the pupil plane of the projection optical system The information about the wavelength of the light of the light source, the information about the aperture of one side of the image plane on the side of the image plane, and the target pattern formed on the substrate are calculated in the projection optical system. An aerial image formed on the image plane; a first two-dimensional image extraction step of extracting the two-dimensional image from the aerial image calculated in the first aerial image calculation step; the first main pattern determining step, according to the first The two-dimensional image extracted by the two-dimensional image extraction step determines the main pattern of the mask; the first peak portion is extracted from the first The aerial image calculated in the aerial image calculation step extracts a peak portion, the light portion is in a peak region except that the main -61 - 200931290 pattern is projected on the image plane; the first auxiliary pattern a determining step of determining an auxiliary pattern according to the light intensity of the peak portion extracted by the first peak portion extraction step; and a second aerial image calculation step according to a pattern and the illumination optical system at the projection optical system The information relating to the distribution of the intensity of the light formed on the pupil plane, the information relating to the wavelength of the light from the source, and the correlation between the apertures on the image plane side of the projection optical system Information, calculating an aerial image formed on the image plane of the projection optical system; the pattern includes the main pattern determined in the first main pattern determining step and within the first auxiliary pattern determining step Determining the auxiliary pattern, the auxiliary pattern is inserted into a portion of the mask corresponding to the peak portion extracted in the first peak portion extraction step; a dimension image extraction step of extracting a two-dimensional image from the aerial image calculated in the second aerial image calculation step; a second main pattern determining step, according to the second two-dimensional image extraction step G a dimensional image that determines a main pattern of the mask; a second peak portion extraction step that extracts a peak portion from the aerial image calculated in the second aerial image calculation step; the peak portion is projected on the main pattern The light intensity is outside the region of the image plane, and the second auxiliary pattern determining step determines an auxiliary pattern according to the light intensity of the peak portion extracted by the second peak portion extraction step; Inserting the auxiliary pattern determined by the second auxiliary pattern determining step into a portion of the mask, the portion corresponding to the peak portion extracted in the second peak portion extraction step, thereby generating the auxiliary The pattern-62-200931290 and the pattern data of the main pattern determined in the second main pattern determining step are used as the material of the mask. -63-