TWI231358B - Determination of center of focus by parameter variability analysis - Google Patents

Abstract

Methods for the determination of center of focus and process control for a lithographic tool. Diffraction signatures are obtained from a plurality of diffraction structures located within multiple different focus setting fields. Variability of diffraction signatures with each field are determined, by direct analysis or comparison to library. The variation or uniformity in the chosen feature for all, or a subset of, the diffraction structure in each field is calculated. The variation or uniformity may be represented by any measure, including the standard deviation or the range of values of the chosen feature or the variability or uniformity of the diffraction signatures themselves, such as by RMS difference or intensity range. The method may be used for process control and monitoring of focus drift by determining intra-field variation of diffraction signatures of multiple diffraction signatures of multiple diffraction structures in a series of wafers.

Description

1231358

[Technical Field] In the age of 丨 = ^, the US Patent Provisional Application No. 6 0/4 62,353, was declared: the decision of the focal length center in the Hai version printing application device (such as Lithographic Applications). The patent benefits of α, ^ > and their detailed items included in April 2003 are hereby incorporated by reference. This booklet involves the analysis of the dimensional changes of the diffractive structures at different positions on a wafer. To determine the parameters of engraving application parameters, including the determination of the focal length center in the engraving application device, such as light Wang Li, resistive stencil printing wafer processing department, and methods of using these decisions to carry out procedures and quality control. [Prior art] ^ Please note that the following discussion involves some authors and publication years, and because they are the latest publication date, some publications are not considered to be past technology in comparison with the present invention. These publications are discussed here for a more complete background ' and should not be interpreted as a prior art for patent granting purposes. Physical Solutions Engraving has a variety of useful applications in the semiconductor, optical, and related industries. Engraving is used to manufacture semiconductor devices such as integrated circuits produced on wafers, as well as flat panel displays, magnetic disk read heads 4 and the like. In some applications, lithography is a method of spatially regulating light to transfer patterns from a mask or scribe to a resist on a substrate. The resist layer is thus developed, the exposed pattern is etched away (positive resist) or pre-retained (negative resist), and a three-dimensional space is formed in the resist layer.

Page 5 5. Description of the invention (2): = type. However, in addition to this photoresist type engraving, type 1 type engraving is used. In a certain type of engraving printing technique, # 导 J uses a wafer stepper; its typical examples include a scale-down = drop: = mirror stage, an optical score board marking table, cassette crystal :: working edge . Modern stepper equipment resists positive or negative methods: both use the original step-and-repeat form or step-and-scan exposure and focus and distance to determine the quality of the developed image type = type stereotypes The resist layer. # 光度 Deciding the average energy of this figure 2 is set by the lighting time and intensity. Facing—The degree of change in the image at the focus is reduced. The local variations in the exposure and focal length of the focal plane, relative to the position of the imaging system in the plane of the focal point, can be caused by changes in the thickness of the resist and shifts in the focal length of the lithographic tools. The stereotypes should be monitored to determine the range. Control of exposure and focal length is particularly important where stereotyped sub-micron lines are used. & Various methods and equipment have been used to determine the focal length of steppers and brushes. Scanning electron microscope (SM) and its sea have been used. However, when the scanning electron microscope can resolve below 1 micron (micron), it can be operated by a high-vacuum chamber: the two procedures will be expensive, drift wood, and F will be relatively slow. It is difficult to automate it, and there are also its industrialists. Illumination light boxes and an etch method all describe the form. For example, it is determined by using the focal length per unit of image. Degree, substrate. The resulting map is allowed to be used to generate a stencil-like I-like device with a shape of 0 · $ use / kind. Optics 1231358 5. Description of the invention (3) Microscopes can be used, but they do not have the analytical ability to process sub-micron structures. Other methods include the development of specialized targets and test masks, such as those disclosed in U.S. Patent Nos. 5,712,707, 5,953,128, and 6,088,113. The gold plating error method is also known, such as disclosed in US Patent Nos. 5, 9 5 2, 1 32. However, these methods still require scanning electron microscopes, light microscopes, or similar direct measurement tools. Various scatterometers and related devices and measurement methods have been used to describe the microstructure characteristics of the following facility materials: microelectronics and optoelectronic semiconductor materials, computer hard disks, optical disks, micro-polished optical parts, and other lateral dimensions in Material materials in the range of tens of microns to less than one tenth of a micron. For example, the CDS 2000 optical scatterometer is manufactured and sold by Accent Optical Technologies, inc. It is a fully automatic non-destructive critical dimension (CD) measurement and profile analysis system. Part of it is publicly disclosed in U.S. Paten1:

No.) 5, 70 3, 6 9 2. This device can repeatedly analyze key dimensions smaller than 1 nanometer, and simultaneously measure the profile profile and perform the estimation of lamination thickness. This device monitors generally diffracted light, which may include, but is not limited to, the relationship between the early diffraction order of the intensity j and the incident angle of the illumination beam. Changes in the light intensity of the zeroth or reflection order of the sample and higher diffraction orders can be monitored in this way, and the information provided is used to determine the properties of the target of the irradiated sample. Because the process used to make the sample target determines the nature of the sample target, this information is also used for indirect monitoring of the process. This method is described in the literature report on semiconductor processing procedures. There are some materials that teach optical scatterometer analysis methods and devices, including

1231358 V. Description of the invention (4) Contains the following US patent numbers 4, 7 1 0, 64 2, 5, 1 64, 790, 5, 241, 36 9, 5, 7 0 3, 6 9 2, 5, 8 Proposed by 67, 276, 5, 88 9, 59 3, 5, 912, 741, and 6, 1 0 0, 9 8 5. Another technique to determine the best focal length is to use a reticle specially designed based on phase shifting technology (r. Edwards, P. Ackmann, Fisher) '' Characteristic description of the consistency and accuracy of autofocus on AS ML stepper using phase shift focus monitoring reticle ", SPIE Conference Proceedings 1997, Vol. 30 51, Vol. 448-4 55). When Appearance characteristics are when the image is taken farther away from the best focal length, the image printed by the reticle becomes asymmetric, and its image is displaced laterally. The analysis of these images can be based on the image Tools for weights and measures, such as those using the ^ key gold measurement method. Another technique for determining the best focal length is the line_shortening technique, also known as "schrntzUmetry" (cp Ausschnitt, ME UgUS) "," Observing Forests and Finding Trees: Critical Dimensions (CDs) Control a New Way. Path ", SPIE Conference Proceedings 1998, Vol. 3332, 21 2-2 p. 20). This method uses relatively large C D η materials &, 1 to. ,, 11 to (3 city meters) line / space array matching: placed next to each other. The structure is printed exactly through the focal length and / or agent mark, and the lines are shortened and large by themselves. The measurement of this space can be made using image tools, such as gold-plated measurement methods. …, The degree of local accuracy, which is used by Bogong to determine the best focal length. The Bossung drawing method is also used. One of the technologies used is, for example, CD-SEM or optical scatterometer. Dimensions (CD) are measured outside a selected focal length

1231358 V. Description of the invention (5) The key rule for measuring the appearance is a parabolic curve and the zero point, and the forest edge diagram is identified. The actual production process of the Parson method is related to the best focal length of the production process, and it is difficult to affect the CD-SEM when it is used in combination. Therefore, the optical scatterometer method is produced. Method Θ In a method of ordering some laser light types, the surrounding phase of the irradiated variable phase plate generates a certain diffraction phase. When using optics, it is possible to diffract the incident source energy source or other combination or the target's diffraction inch. The advantages of Jiajiao's key size are not always used to automate its measurement of large deviations and their relationship to a certain continuous range of beam sources, which can be selected individually, and a wide range of results will be used. Trend trend matching distance. One of these songs is that in addition to the quantification, it is sound and implemented in a way. Small may suffer the result ° Usually parabolic. Set the slope of the curve to the line known as the best focal length outside the system. However, this method makes it difficult to determine the most. Furthermore, when the method is used with a device that changes the angle of the side wall surface of a line, the device may use a single light source with a known wavelength for various operators, and its range of incident angles is optical. In addition, the filter angle and the filter angle are interactively changed around the radiation source of the target to identify the target light. The incident light, which is also a component, may also cause the light sheet to change in a certain range φ dry area and rotate. Switching to use, all 0 changes. Another option is to use different spectrums of incident light and its angle of incidence is known. This phase can be adjusted within the polarization state of different polarized light, or it can be determined by any of these methods. The angle of incidence Θ . The source has an optical element that can be detected by Θ. The optical elements detected by the optical detector from S to P cause the light source to have different devices with the standard phase. It is known that one can use one at another wavelength range. for sure. This is the part formed by the filter and filter. In addition, it is also possible to vote for the light and one of the samples.

Page 9 1231358

In addition to the optical scatterometer device, there are other devices and methods that can determine the zero or higher order diffraction identification mark, which uses a kind of reflected or penetrated a periodic structure, and is intercepted by a detector . In addition to I optical scatterometers, these other devices and methods include ellipsometers and light reflection meters. In addition, it is further known that other radiation sources, such as X-rays, may be used to obtain non-light-based diffraction identification marks. Tian has known a variety of specimen targets that are well known in this art. A kind of single-marked light beam used for hanging is a diffraction grid bridge, which is essentially a series of periodic lines with a representative width-to-space ratio of 5 between i and 1: 3, although There are other ratios. A typical diffraction grating, for example, a ratio of 1: 3, may have a line width of 100nm and a spacing of 301nm, and the total spacing (width plus spacing) is 400nm. M / the width and pitch are a function of the resolution of the engraving printing program; therefore, if the engraving printing sequence is small, the width and pitch can be reduced equally. Diffraction techniques can take advantage of any feasible width and spacing, including those that are inherently smaller than the width and spacing commonly used today. Bi-periodic and other multi-periodic structures are also known, such as those disclosed in US Patent Application Publication US 2002/0131055, published on September 19, 2002. Second-degree spatial grids or structures are also known, including those disclosed publicly in U.S. Patent No. 6,42 9,93 0. Therefore, the diffractive structure may have more than one cycle, or it may be composed of elements other than lines and spaces, such as holes, squares, rods, and so on. It is further known that diffractions generated by a non-periodic structure, such as isolated feature appearances or series of feature appearances, can also be used as the methods discussed herein or the scope of patent applications.

1231358 V. Description of the invention (7) Diffraction structure In the grain. Known (or exposure field) made by brushing method, dose. At the same time, the key dimensions obtained by comparing different model database comparisons with the model have been combined. This Parkson system is based on a technique known today. Each diffraction is determined, for example, by using the position of the focal length of the unknown, and the key scale focal length is plotted to know that the structure can be used at the same focal length in the model, and the focal length of the structure can be used. As discussed above, typically a single wafer can be identified by diffraction with no setting or scattering measurement. This fact was deduced, and the results were deduced to the critics, and they were distributed on the same focal length, different exposure techniques, and inter-diffraction indices. Parabola has obvious stereotyped light setting or emission structure of multiple grains on the wafer, and its theoretical identification mark state will be limited by this curve. US Patent Nos. 6 '42 9, 93 0 and 6, 6 0 6 1 52, as the same inventor of the present application, teaches a parameter measurement method related to a stencil printing device, which uses a step of providing a substrate whose substrate contains a large number of Diffraction grid formed on a substrate by a stencil printing process, the diffraction grid includes a plurality of spaced-apart elements; a radiation source-based tool measures the diffraction of at least three of the plurality of diffraction grids Identification marks; and measuring the differences between these diffraction identification marks to determine the so-called engraving printing device—the desired parameter. In this method, the substrate may include a wafer. In addition, the method may include a large number of diffraction gratings, which are used for engraving printing devices with different known focal length settings; and determine the diffraction grid book of two adjacent focal length settings, | The difference between the marks is smaller than the differences between the diffraction recognition marks between the diffraction gratings set by other adjacent focal lengths; this parameter is the focal length of the engraving printing device.

1231358 V. Description of invention (8) Heart. That is, the inventors of the international patents among other marks taught that they would compare the diffraction library; the closest key to the diffraction key size (CD) made by this section could be replaced by the cross section. This section can be centered on the adjacent focal plane itself. At the most important position on the section [Content] The method of the present invention includes the inclusion of numerous numbers and numerous numbers in the field of the multiple numbers in the general base; and the same applies to the diffraction of each step. The measurement method is based on the model data. The diffractive structure is obtained by changing the focal length. For example, the obtained focal length may be a non-geometric ratio, phase, or the point where the focal length is zero. At the focal length, the difference between adjacent focal length steps will reach a minimum. Application PCT / US 02/3 239 4, as in the present application, the cross-sections of the diffraction identification marks and the structure of the diffraction structure measured by a parameter structure related to the engraving printing device are based on the cross-sections of the model and the model. Decided. This method is repeated for the structure. Substantially changed parameters, side wall surfaces, or thickness of the resist are calculated and calculated; these parameters may be the numerical differences in the cross-sectional area of each structure in the step area of the largest cross-sectional area step obtained from the area, volume, and a focal length trend. Illustration. In these examples, the curve 'usually fits into a parabola. In the latter case, the minimum or maximum value of the product is not needed for the curve-to-fit combination. : The type of rigidity and the relevant parameters of the one-plate printing equipment = a kind of substrate, preferably a semiconductor wafer, the steps, fields, and fields have been exposed at different focal lengths. Use this engraving to print ti = achievement; and use a radiation source as the basis and name iiii for each diffraction and structure to measure its 'diffraction identification mark s /', which is composed of a large number of diffractive phases in the field.

Page 121231358 V. Description of the invention (9) Obtain the variability of the projection identification mark; 1 and compare with these fields = to determine that a desired parameter change of the engraving printing equipment may be monocyclic, double -Seoul, multi-period "biological" or non-periodic structures, including grids. The Haijintian radiation source as the basic tool preferably includes a kind of light margin light, 7? ≫ 7 ί :: tools, preferably containing-or more incident laser light ί, ί =…; The diffraction identification mark formed by the entire measurement angle ^ ^ = intentionally includes a scatterometer analyzed by angle-division. Another light ΐ 隹 ;; Ϊ: a kind of wide-spectrum light source, an optical system will detect light emission at a wavelength of 波长, and the surrounding wavelengths, and a detector formed by ^ two remaining wavelengths Diffraction identification mark. The tool can include 3 sources of incident light, whose components are those that change the amplitude and phase of the S and P polarizations, an optical system that focuses the light and emits light at a certain range of incidence 2 2 and a detector The diffraction identification mark formed by detection is measured according to the above method. The diffraction identification mark can be arbitrarily planted with a wide-spectrum (: Ge, also selected, single-wavelength) radiation source as a basic tool for phase measurement. A fixed angle and a variable incident angle produce a variable z-scanning angle F to operate; or use a separate multi-wave phase measurement. The diffraction identification marks can be any kind of reflective, conductive, reflective, or ancient diffraction identification marks' and can optionally be ordinary light scattering. Measurements. The desired parameter is preferably the focal length center,

1231358, and preferably the domain association step can be arbitrarily surrounded by its measured number of identification marks, such as error. The determination of the identification distance, the ^ mark determines a theoretical diffraction identification mark, the number of diffraction appearances is the largest, the cross-section body diffraction structure intentionally includes the diffractive structure, and the standard of the nose is included. There are many places, and the common denomination is determined by the value of the most desired parameter with the diffraction identification mark. Including the measurement of the diffraction identification system for each field, which is obtained by a large number of numbers in the field, or to calculate the root mean square of the diffraction identification system (root includes The information database contains the theoretical diffraction recognition step package information database for each measurement; the best fit 4 correlation; the structure is relatively good, or two or every unit is associated with the unit variation (s于 知道 数 演 # Most diffractive structures 5. Description of the invention (10) is a small dose field of degeneration. The intensity of the step marker is determined. The diffractive statistic of the diffraction pattern is measured. Mean square) Optional. The theoretical diffraction obtained by the diffraction identification of the best-adapted measurement of the diffraction position in the field degenerates in the field. The rationale for determining the selection of a cross-sectional area and the adaptation step can be measured with the measurement in the field Alternatively, the selected appearance scale method can be used to determine the effect of different doses set at the focal length. Knowing a selected appearance with different focal length settings and the selected appearance key size (CD) associated with each field, or should provide a more theoretical The field size of the external dimensions is measured with a number of statistical features of the selected external dimensions. Measure the different focal length setting structures of the tandard dev iat, and the measurement structure can be set at will. These sets are marked; the size and the measurement are related to The size can be selected by the diffraction identification product. The number lies in the range, the value, such as ion) ° and its known dose pair contains the

Page 14

1231358 V. Description of the invention (11) Different dose settings vary. The diffractive structure of the above method may optionally include a latent image diffractive structure, and the substrate may optionally have not yet undergone a developing process. The invention further relates to a program control method for a focal length center in a printing plate printing apparatus. The method includes a step of measuring the focal length center of the printing plate printing apparatus according to the above method; and adjusting the focal length setting of the printing plate printing apparatus to The focal length center of the measurement can be freely used by a computer-based control system or an autofocus control system. At least one of the input systems of the autofocus control system includes a measurement value related to the minimum chirp. The variability of the measurement diffraction identification mark obtained from a large number of diffractive structures located in a selected field is best for time measurement, and the selected field is preferably one previously measured at the center of the focal length. Preferably, the focus of the engraving printing equipment will be adjusted if t 忒 is acceptable and the diamondness exceeds a predetermined control limit. The invention is also a engravable comparison of a engraved printing device obtained by measuring a diffractive structure around a number of tools in a number of diffraction field domains with a series of structures and measuring the diffraction structure with the wafer. The method of program control in the plate printing equipment is structured and the exposure steps of the plate printing equipment on a certain 3 wafers are mostly in a certain field and the variable number associated with the measurement winding is obtained with a kind of radiation identification mark. The method has at least one 'denaturation. It can be used to identify the signature of each shot to control it. It is better to include other parameters. The degeneration is best to determine the variability of the wafers on the series of wafers. Based on experience or emission source, each diffraction is associated with a large number and the device is compared to adjust the circle.

1231358 V. Description of the invention (12) The determined variability limit is compared. Focal length or dose. Such diffractive structural, multi-periodic, or non-periodic circles preferably include semiconductor wafers. The radiation source is the basic tool. It is best to include an OR system to focus the laser beam and scan a detector to detect the entire measurement. The tool can optionally include the user, the tool includes an incident to focus the light and to enter a certain range to detect the shape of the entire measurement wavelength. It randomly includes the amplitude and phase of an incident light source, an optical incident Phase luminescence, as well as a kind of distance detection. Measure the broad spectrum (or choose another method) with phase measurement according to the above method, and operate at a certain degree or a variable scanning angle F; Ik is a basic tool for wavelength radiation source. A reflective, orderly-diffractive diffraction identification mark, and a certain amount of measured value. The desired Lv Hai $ / | > With, a desired parameter preferably contains sexual dagger 3 early periodic, double- The structure of the period f is like a grid. These crystals are best & «a kind of incident laser beam source based on a light source, a light incident through a range of incident angles, and a diffraction formed by the angle A scatterometer to identify the division of the scale. See also the spectrum light source, an optical system to emit light at a wavelength, and a diffraction identification mark formed by the detector. The tool can be composed of changes in the S and p poles. The focusing system focuses the light and uses a certain range of detectors to detect the formed diffraction identification mark. The identification mark can optionally include a fixed angle based on a single wavelength) radiation source, a variable angle of incidence, or Choose one more Divided multi-phase measurement. The diffraction identification marks are guided, reflective, or higher. Randomly general light scattering or diffraction parameters are preferably at the center of the focal length, or at random.

Page 16! 231358

V. Description of Invention (13)

Ground is a dose, and is preferably determined by a measurement of a desired parameter associated with a field having the smallest variability of the diffraction identifier. This decision step may optionally include measuring the intensity range of the diffraction identification mark for each wafer, and the identification mark is obtained from a plurality of diffraction identification marks measured on the wafer located in the field, or otherwise The selected person is the measurement value of the ancient ten different variability statistics, such as the root mean square error of the diffraction identification standard. ° Alternatively, the theoretical diffraction identification mark generated by this decision step may be used. The diffraction identification mark measured by the data determines one; the best fit theory diffraction identification is compared with the measured diffraction identification mark and the variability of the external appearance size on the wafer in the field. . The selected external one can be selected as a theoretical diffraction product of a cross-sectional area and an adaptation of a diffraction identification mark. The decision step may optionally include a range of a large number of diffractions on the wafer within the field, or alternatively include values such as the deviation of the selected appearance dimension). The diffractive structure can be constructed, and the wafer can optionally be provided with relevant parameters of the device without the primary objective of the present invention, and the unused light includes providing a theoretical diffraction junction library; The theoretical theoretical diffraction identification of the suitable identification is associated with a selected external dimension, and for each wafer, the selected external dimension associated with the majority of the diffractive structures is preferably a critical dimension, or area volume, or provides a theoretical diffraction structure. Multiply two or more physical dimensions by measuring each wafer's selected physical dimensions associated with the structure. Calculate the statistic of this variability. Measure the quasi-deviation (standard optionally contains the latent image. Diffraction. A development program. . Method for measuring dry, scanning electron microscope for engraving and printing

Page 17 1231358 V. Description of the invention (14) (SEM) or a microscopy-like measurement tool. Another object of the present invention is to provide a method for determining the focal length center of a block printing device by diffractive identification marks by analyzing the best-fit theory of a series of fields of different focal distance diffraction structures, the diffraction structure including but not Limited to the diffraction grid; and the inter-field variability of the best-fit structure is used to determine the focal length center. Another object of the present invention is to provide a method for determining or measuring the relevant parameters of the J-type J printing equipment, including the focal length center, by using reflective or conductive diffraction methods to obtain some of the same in different focal length fields. The diffraction identification mark of the focal length structure; and determines its internal-field variation between such diffraction identification marks, or a certain derivation from a theoretical model that provides the best fit theory diffraction identification marks Internal of parameters-field variation. Another object of the present invention is to provide a method for determining or measuring the relevant parameters of a block printing device, including the focal length center, obtained by using any method to generate a diffraction identification mark—diffraction identification mark; the method includes but not Limited to reflected or transmitted angle-divider, variable wavelength, variable phase, variable polarization state or variable azimuth diffraction, or a combination thereof; its identification mark is any diffracted or scattered light The zeroth order or the diffraction order of reflection or higher. Another object of the present invention is to provide a method for determining or measuring a moment = relevant parameters of printing equipment, which are used to identify diffraction levels or diffraction levels of any order of different focal length fields in a wafer or other substrate. Higher-order diffraction, whether positive or negative, or any diffraction

in page 18 1231358 V. Description of the invention (15) or scattered light. The first advantage of the present invention is that it allows the measurement of parameters related to engraving printing equipment without the use of optical, scanning electron microscope (SEM) or similar microscopy tools. Another advantage of the present invention is that it provides a method and equipment that allow the results of engraving printing equipment, such as a stepper, to be included in a shorter period of time, at a lower cost than the commonly known methods, including the focal length. center. Regarding the other objects, advantages and new features of the present invention, as well as more applications, etc., some will be put forward in the following detailed description, accompanied by related illustrations; some will be obvious to those skilled in the art who have passed the following checks. Or it can be learned by practicing the present invention in the field. The objects and advantages of the present invention may be achieved and obtained by means of means and combinations thereof, particularly pointed out in the scope of the attached patent application. [Implementation method] The present invention provides a method and a device for measuring a related parameter 5 of a printing plate printing device, and in a preferred embodiment 5 it is determining a focal length center of a printing plate printing device. In the photoresist development step of the wafer processing procedure, it is critical and critical to determine the focal length center for a fixed dose. Furthermore, variations in dosage can increase the difficulty of determining this center. Lenses used for engraving tools have a very limited depth of focus, so maximum accuracy is required. When the lens is at a positive focal length, a sharp-cut photoresist image will be produced, and the lack of correct focus will result in a photoresist appearance that is erroneously developed, and usually a relatively crude procedure is produced. At the center of the focal length, or at the optimal focal length, the repeatability and stability of the program are improved. One that determines the center of focus

Page 19 1231358 V. Description of the invention (16) The method is disclosed in this application example using variability analysis; in short, the focal length center is determined by the field consistency of the diffractive structure. The invention further includes manufacturing process monitoring based on the consistency of the diffraction identification marks. Before further describing the present invention, the following definitions are given. When used in the full range of specifications and patent applications, variability or variation means the extent of the value of a quantity or parameter, which is measured or calculated for each desired group item Between, including but not limited to diffractive structures, which are different from each other. Variability and degree of change are the opposite of consistency, and these nouns can be used to determine ^ =, measure 1, calculate, or compare their variability, so they are consistent with decision, measure, difference, or comparison Are synonymous. For example, I Chu Γ 3 "quotability" is synonymous with the term "maximum consistency", and in February, all specifications and patent applications can be interchanged with a engraving such as a mask. . This includes printing, but it is also known as the optically engraved master image. Also known as the sequence, whether or not the wafer is made of negative resist material is printed. Device refers to any device that uses an image including It is common to include wafers with other stencil printers as masks or more. Further processing may be used to transfer a pattern to a substrate. In the ‘optical method reticle, the transformation is called anti money anti # coating agent system, for example. For example, a positive resist may be inserted into the substrate, for example, photoresist engraving. Photoresist engraving is used to print a circuit pattern on a wafer. During this process, the agent will be coated on the circuit and used as required, that is, as long as it is installed and baked. Positive resist or usually insoluble in is used as

1231358 ------- V. Description of the invention (17) The anticorrosive developer's chemical negative resist is usually dissolved in: used: two negative fouling agents that become soluble when f-line is usually dissolved in the used For two = deer, become soluble when the line. Light becomes insoluble when light: "Will the chemical, but expose the ::: area and exclude the other parts? = Layer selective exposure can produce selective exposure in the resist layer by the circuit or other social agents There are two types of law system: In the first-grade engraving printing, the choice is: to direct the light to the light; to cover: the typical representative method layer of the mask. The image of Zhu Detong is projected onto the resist film. The known t-printing printing device includes a stepper and a scanner, or other structures. 乂 A machine or a wafer scanner; it is used to connect a circuit. A typical generation: the image is covered by a light The mask mold is projected onto the resist-coated wafer light source, θ ^^^ advancing machine includes a reduced lens and light emitter, laser excitation and operation work ^ Mirror building, reticle optical marking mirror stage, Cassette wafer cassettes using stepping ~ 1 Cao ... Stepping machine adopts both positive resist method and negative resist method, and it is used in a complex form or a step-scan form, or a combination printing device for both. T ^ Ming is implemented by a series of diffractive structures to clarify, _ Yan Yi is made on a wafer or other substrate. In the simplest way In the production of the formula, the projective structure is any structure or image that is hand-made with a engraving printing tool. This project produces a spatial change in the refractive index of the incident illumination. The change in level may be due to physical differences or chemical differences. For example, changes in photoresistance or other engraving printing methods are generally used for engraving. Some materials with a refractive index and Air binding 'is like an optical diffraction grid of s, or a substance is bound to other substances

1231358 V. Description of the invention (18) One of the differences includes the wafer having an optical resist exposure order, such as a grid, in which the resist has not been developed. Here #Etchant is still present, but the developed part has a different refractive index, ie, a knife, resulting in a diffractive structure that consists of a periodic change in resist uniformity. This periodic change is obtained by the periodicity of the knotted knife. A diffractive structure may have a single = double-circle ㈣, or may be multi-periodic, or may not; :: 'This is also possible because of the diffraction of a single feature appearance ^ 5 ordinary diffraction lattice composed of a series of parallel lines Grid,: Du ί 2 eyes and other three-dimensional arrangement of the structure, in which X: "〃 periodicity 'and neither in the X direction or Y direction; two: = diffractive structure with periodicity can be, for example, ^ 4, or a relatively large single rectangular structure, or ,: equipment structure. Similarly, the above-mentioned diffractive structure is being issued: Second, the complete period (the structure is repeated multiple times (one more time)), And it may only appear repeatedly a few times (for example, a structure includes a photoresist grid, a surname: Yi = grid, and other people in this technical field ::,,: shooting grid It typically has a straight line width and an interval ratio of 1/1: 3, although other ratios are also useful. — = Grid, for example, a ratio of 1: 3 may have a distance of 1◦n nanometer line J The width and spacing can be re-obvious, and the resolution of the engraving printing device No. 4 In the implementation aspect of the present invention, there is an f-emission structure. In the case of all non-developing parts, the refractive index: or the chemical cycle can be any. Therefore, there are also t-direction and Y-side periodicity. It is a simple or more complex beam which can be 10 or three times.) Folding grille, thumb.-About 1: The apparent winding and 40 nanometers are determined to produce a 1231358. 5. Description of the invention (19) Diffraction identification marks. Diffraction identification marks can be produced by any of the following instruments, such as optical scatterometers, ellipsometers, or light reflection meters; optical techniques such as scatterometers, interferometry, Polarimetric measurement. Reflectometry, spectroscopic ellipsometric measurement, or spectroscopic reflection measurement, and use any technique, such as angle or spectrum analysis, etc. Any device or device that uses a radiation source as the base radiation source is The base may be a diffractive identification line, which is the angle Θ emitted by the scatterometer. The tool can be a tool for generating rays. The tools of the light can be reflected by external standards. The typical identification of radiography, such as the source of light, such as X-ray reflection mode, a single unit in this diffraction identification has been classified as a type of user, a kind of stone, and the watch is adopted as the basic line source. The reflection of the source can indicate the wavelength change. The reflection angle source 'may' means that the light of the perimeter is constant. The emission phase of the variable phase, the light source with a detection polarization is also a human component or a P to S component. The light source and the diffractive structure are relative to each other or in a continuous range within a continuous range. For the method of incidence and incidence, some sources of incident angles of the laser beam are used, and the Θ can choose to use a certain polarization around the diffraction. In a range, or these different Θ. The range of incident light is maintained in place systematically. The variability is adjusted from S to p within Φ and selected as a device. In another method, it is a tool with a certain wavelength range; but in a certain implementation, the ray can be caused by the user with a light source of an angle. The green graph describes that each of them randomly chooses a wide-incidence source to emit, and the light source is also known by the human detector, which makes it possible to rotate around the diffractive light source. There are three changes in the column. In the example of the radiation source, the radiation source is divided into two categories, such as the photometric division and the incident identification target. It is known that the incident angle of light with a different spectrum is different, and its measurement is caused by a certain range.

Page 23 1231358 V. Description of the invention (20) Knowing the diffraction identification of the known diffraction structure ^ $ 志. Generally, the detected light intensity is plotted for at least one variable parameter, such as the angle of incidence Θ, the wavelength of the incident light, the phase of the incident light, the scanning angle Φ, or the like. Diffraction identification can express the diffraction order of zeroth order or reflection, or it can represent any higher diffraction order 'or the measured value of the diffraction and bite scattering of general light rays. It is also possible or foreseeable that it is a conductive mode that can be used to generate a certain diffraction identification mark, such as using an X-ray radiation source as a component of which the radiation source is a basic tool. In one embodiment of the present invention, 'a theoretical information library that generates diffractive structures and corresponding theoretical diffraction identification targets', and theoretical diffraction identification marks based on the theoretical diffractive structure and measured diffraction Identification mark comparison. This can be done by any of different methods. In a certain method, the actual information base of a theoretical output signal is generated based on the specified parameters of the variables. This information database may be generated before the actual measurement diffraction identification mark, or it may be generated in the procedure where the measurement diffraction identification mark is adapted to the theoretical diffraction identification mark. Therefore, if used for this purpose, a theoretical information library includes a theoretical information database that is generated independently of the measurement diffraction identification marks, and a theory based on the geometric shape of the measurement structure "best guess π and measurement diffraction identification mark products The theoretical information database generated from the calculation results; and iterative comparison with the changed parameter structure to determine the best fit. The diffraction identification can show that the theoretical information database can also be generated empirically, for example, by Collect diffraction and structure identification marks, whose structure has dimensions measured in another way. The information base can be freely reduced by removing certain signals. Its No. 5 system can pass other signals in the reference group. Within the difference and correctly expressed. The information

Page 24 1231358 V. Description of the invention (21) The index of the library can be generated, which is similarly by associating each identification mark with one or more index functions, and then sorting the indexes according to the degree of the correlation. The construction or generation of this type of plastic information database and its optimization methods are well known in terms of technology. In a certain method, a rigorous theoretical model based on Maxwei's equation is used to calculate the predictive characteristics of optical signals used for diffractive structures, such as diffraction identification marks, as diffractive structures. Argument function. In this process, a set of experimental values with this diffractive structure parameter is selected. Then, based on these values, a computer—a model that can represent the parameters of the diffractive structure—including its optical material and geometry was established. The interaction between the diffractive structure and the luminous radiation is simulated numerically to calculate a predicted diffraction identification mark. Any kind of trial optimization algorithm can be used to adjust the parameter values of the diffraction identification mark, and iterate through the program to minimize the difference between the measured and predicted diffraction identification marks. In order to get the best fit. The United States Federal Publication Patent Application US 2 0 0 2/0 046 0 0 8 discloses a database method for structural identification, and the United States Federal Publication Patent Application No. US 2000/0038 1 96 discloses another a way. Similarly, the U.S. federal issued patent application number US 2 0 0 2/0 1 3578 3 discloses various methods of using the theoretical information database, as the federal issued patent application number US 2 0 02/0 0 38 1 9 6 — kind. The grating, or the diffractive structural parameters that may be used in the theoretical information database, include any parameters that can be used as a model. The included elements are, for example, the critical dimensions of the part and 7 or the top. The degree of drought, such as a line, Height or thickness of the post or other structure

Page 25, J231358 V. Description of the invention (22) Degree:-The total height of the range defined by the diffraction identification mark is circular or its shape of a structure, such as a rectangle, trapezoid, triangle, other geometric shape or area The radius of curvature at the bottom and / or top of the range. Side wall surface. • The period of a grid. • The width of the line or other structure. The average value, such as the critical dimension at a given position, the weight value of the contribution between the structure and the substrate, or the like. The thickness of the film and the optical properties of the material are usually not relevant to the information model of the dance distance. Causes noticeable changes. Uninteresting ::: Theories The tragic configuration and the choice of theoretical models depend on the change of distance ':: yi, some changes in properties. This may be very fast, and can cause this. The film thickness and optics with the determined cross-section are often equal to: the library may include the same periodicity in the cross-section ^ may be simple, or may be Complicated. For example, the contribution of the area of the thumb cross section grid and the cross section of the film is given by n &k; considering the introduction of films and patterns. σ μ. This theoretical model may also be used in one embodiment of the present invention. The cross-sectional area of Zeng Zeng, the identification mark of which is compared with the measurement ^: The theoretical diffraction identification mark, the comparison of the 5 marks

1231358 V. Description of the invention (23) Shiyi __ & Zhejiang area is the best matching theory diffraction identification mark obtained in the present invention. Office, 44- ^ ^ ^ β-within the meaning of the parameters of a diffractive structure ’the best. Product of the best-fit structure. In one embodiment, the cross-sectional area is the area of ▲ ▲, such as the product of the critical dimension and the height. In another embodiment, the cross-sectional area is a cross-sectional volume, such as the product of key dimensions, heights, and structural shapes. However, if applied to this, the cross-sectional area need not be a shape defined in any way; that is, the cross-sectional area may be the product of any two or more diffractive structure parameters, including but not limited to the above and By. In an embodiment, the cross-sectional area includes a critical dimension and a parameter of another diffractive structure. If applied to this, the so-called product of at least two diffractions and structural parameters is the mathematical operation or operation of the at least two parameters, including, but not limited to, a mathematical operation including multiplication, and at least a second one is optional and optional. computation. "A wide variety of theoretical models can be used to determine the profile contour area, the area of the cross-section grid area of 3, such as 5 'rectangular grid is defined by this formula: section area = Η ·? (Ι) where Η is the grille height; # & „is the width of the thumb in order to improve the pot fine. For determining the focal length center of the stepper. One such model is-chess degree, and a more detailed theoretical model may be used. The shape of the trapezoidal grid cross section is added to the size of the side wall angle. Determine the cross-sectional area = H · The formula is where Η is the high-pitched tan A) (2) plane angle. Other, V complex: 1 width of the bottom of each grid, and A is the complex shape of the trapezoidal side wall, for example-trapezoid has a rounded edge, height 1231358 V. Description of the invention (24) Si or S-shaped outline, or specified by the user Other customized contours may also be used to generate theoretical models. More complex shapes require more complex equations to represent the shape of the grid in cross-sectional area. A second degree spatial structure may also be analyzed in a similar way. For a three-dimensional structure, a measure of the section is the section volume. For example, a simple contact hole model is assumed to be a perfect circular contour in the X-axis and γ-axis directions' and a side wall surface that is not interrupted at a 90-degree angle along the Z-axis, and the section can be viewed as Calculate for a cylinder to produce a section volume. In addition, the non-geometric cross-sections of the unitary elements that are seen in the image may also be used. =, The so-called cross-section may be a certain parameter, such as the critical dimension, and the multiplying m parameter of the 1 external parameter includes, for example, the mean value of the material, the measured value of the angle, the optical properties, and the curvature. It's the way of volume. The break of a certain theoretical diffraction identification mark. The best adaptation or adaptation of the bidders can be studied; 2 Diffraction recognition * formula calculation. Seen in a certain embodiment ;; an adaptation of the known diffraction identification marks to separate and existing diffraction identification marks that are included in the measurement, the theory revolves around the use of a defined beam II; for example = Using various adaptation algorithms, in another embodiment, the most suitable adaptation can be selected. In a certain-theoretical diffraction identification club, :: 1Beijing: within • 'to get before the internal difference is not stored. Even so, the theoretical total number of possible theoretical diffraction identification targets, based on Computation of the flat description of the correlation model-a kind of theoretical diffraction; the theory; the sampling of diffraction identification marks to carry another J. Yuan. In this way, any

Page 28, 1231358 V. Description of the invention (25) Method or technology for confirming or adapting the theoretical diffraction identification mark, no matter how it is obtained, may be used here to determine an adaption or best fit. Match.

The diffractive structure is typically produced on a resist material, which has an opaque and transparent area prepared from a masking mask ', which conforms to the required shape, size and configuration of the desired diffractive structure. A radiation source is applied to one side of the masking mode, whereby the shape and space of the masking mode are projected on the reactance, layer and the resist layer is located on the other side of the masking mode. One or more lenses or other optical systems can be inserted between the hood mold and the reactor, and optionally between the radiation source and the mask. When the exposure layer is exposed to radiation or a sufficient amount of energy is applied to change the resist layer, the image formed in the anti-residue layer is formed. The potential image, generation

The chemical change of the substance of the anti- # layer results in the change of the internal reflection ratio of the resist layer. Therefore, 旎 is used to generate the above-mentioned diffraction identification mark. In one embodiment, a wafer with a latent image on the resist layer may be subjected to post-exposure bake to promote more additional chemical reactions or to dissolve more components in the anti-residue layer. In another embodiment, the resist layer can be developed by a developing process, and the process can be selected as a chemical developing process, in which the scratch of the resist layer is removed, and the portion thus removed can be determined as Utilize positive etch ^ 疋 疋 negative resist. The development process can also be classified as an etching process to create the area or space where the resist is etched, and can optionally be a substrate material, such as other film layers, that is, the resist layer is placed on it. In the method and device of ^ Ϊ Ϊ Ϊ Ming, the true diffracted light tree may be developed ^ or may be developed arbitrarily. Similarly, when the foregoing

1231358 V. Invention description (26) Generally, when describing a common method of generating diffractive structures, any method may be used, including the use of phase shift masking modes, any variety of radiation sources, including electron beam exposure, and Its wait. Focal length is a key parameter in any stencil printing equipment, including a stepper or similar stencil printing device. Focal length and focal depth are functions of the dose, or total amount of radiant energy; and focal length, or the distance from the lens to the target. The resulting image must be good for all points in a given exposure field, resulting in a focusable and usable depth of focus. However, factors other than dose and focal length also affect the focal length and focal length depth; it includes astigmatism, field curvature, lens quality, wafer stage orientation in X- and Y-directions, and so on. A typical production wafer stepper has a resolution from about 0.1 to about 1.5 micrometers, and a usable focal depth from about 0.4 to about 1.5 micrometers. The determination of the focal length center for a certain dose therefore becomes critical in the effective operation of a certain engraving printing device, such as a stepper for the photoresist exposure step during wafer processing. The degree of dose variation increases the difficulty in determining this center. Lenses used in steppers and other engraving printing devices have very limited depth of focus, so maximum accuracy is required. When the lens is at a positive focal length, a sharp-cut photoresist image will be produced, and those without proper focus will cause a non-functional photoresist shape. The focal point center point also significantly improves the repeatability of the procedure. Once the focal length center is known and determined, any of various autofocus systems or structures may be used to decide to differentiate the lenses while the wafers remain constant. These systems include optical methods such as those using reflected light, capacitive methods, and pressure sensing

Page 30 1231358 V. Description of the invention (27) The device method, for example, cannot be used to determine the focus in the hold. In a typical operation, a kind of engraving is used to print the lens aberration and add. The optical system and the diffractive structure with the same focal length beyond the point where the distortion becomes more prominent will pass through the focal point in a priority real field 20 series using each field of the wafer stepper 20 series. The best distance value is best to include If the zero-focus position of the zero-focus position matches the original size, the reed appears at the point of focus, or the focal knee does not match. The displayed sequence of numbers that may change the focal length may also be the same as those shown in Figure 1A. However, these systems and structures only maintain the lens-wafer distance to a certain extent, and the focal length center must be determined periodically. Common devices operate every six hours or so. Distortion increases as it moves outward from the center to the edge. The further away from the focal length, the positive and negative directions, these aberrations are important. In this way, if some lenses are projected by the focal length value, those who are distributed over the entire field of these lenses will exhibit greater variability in the focal projection. In the embodiment, as shown in FIG. 1A, a series of spaced-apart fields are exposed or printed, preferably through a wafer or other substrate at a fixed dose by means of stencil printing. Expose at a different focal length value; these produce a series of focal length centers. In the example shown in Figure A, the amount of focus deviation is represented by the number displayed on each field. The placement is arbitrarily selected, or randomly selected to correspond to a certain physics such as the top position of the anti-residue layer. Since the optimal focal length passes some distance into the resist, the real center of the optimal focal length will most likely not be understood as limiting to the actual number of initial zero focal length positions. The number is greater than + 0. 4 microns or less than -0.4 microns. These are different fields described by which field is projected at which focal length deviation. Is a diffractive structure separated by a large number of 3 0, 3 0, so

1231358 V. Description of the invention (28) For example, it is better to form a diffraction grid, which can form I-P 0 a to cover all fields. FIG. 1B depicts: a certain% domain has 25 diffractive structures 30, 30, separated by 5 X 5; however, any number of social structures v ^^, 口, and 构 are used for this purpose, and their Intervals are not evenly distributed across the 20 domain. Optionally, the field may be composed of two large fields, which are measured on the large grids as discussed below; or may be a combination of the two, in which the field There are some large grids, and each σ

Et you include a large grid of ΛΑ valleys in the grid with different diffractive structures 3 °, 30 ′ complex or periodic structure secret-shape but not enough to constitute a diffractive radiator; and It may be stingy, such as a common, homogeneous, pole, or more complex structure with lines and spaces. It is best that the y = = y, y, y, y, y, y = y, y = y, y = y, y = y, y = y, y = y, y = y, y = y, y = y, y = y, y = y, y = y, y, y, y = y, y, y, y, y = y, y, y, y, y, y, y, y, y, y, y, y, y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, Y, I were specifically made for use in the present invention; however, it was possible that it had a true pass structure, which was being manufactured on each die on the wafer. The basic tool can be used as a light scatterometer measurement method, each of the 2 diffractive structures 30, 30, or its determined sample, that is, measured to obtain a diffraction identification mark at a fixed dose. In the edge-first case, each diffractive structure is compared with the theoretical information base of diffraction identification marks, and each identification mark has a known appearance. A kind of diffractor: the diffraction identification mark of each experiment and a theoretical diffraction identification are found, and a selected appearance size associated with the best-fit theoretical model is assigned to each diffraction Structure 3 0, 3 0 ,. The selected appearance size may include a single shape appearance, such as a critical dimension (measured, for example, in nanometers), or a combination of appearances to form a section (as defined above). This selected aspect size will vary depending on the diffraction

1231358 V. Invention description (29) is different. Preferably, it should be the selected structure of the diffractive structure 30, 30 within each field. Out. In addition to the standard deviation of ^ standard \ outward dimensions, the standard deviation is then used in calculations, including but not limited to ^ = a variety of consistent measurements may also be made (maximum minus minimum) :: ': the overall range of selected appearance dimensions of the domain Degree of change. 4: Other statistical or mathematical methods to represent, such as a library of moments, used to determine breaks ... You can use a chessboard branch 2 with simple edges, or use more complex shapes such as trapezoids, circles, etc. Gaussian or 8-shaped outline shape, or its ΐ =: 疋 outline shape. The theoretical database may also take taxis into account, pretending to be ancient 3 and style. The diffraction identifiers generated in a theoretical manner = the second one, the known characteristics of the target 'are adapted to the experimental data in order to obtain the 1: ordering characteristics of f. More complex models can include the diffraction structure = * such as the optical η and k values, and the model layer below it. The distance gauge measures the sum of the optical path. Second, the standard deviation of the selected appearance size (* diffraction identification ^ / 夂 生 is then plotted for the relative focal length of each field 20, = 2 ^. Although the selected appearance size is the key in this example) Dimensions, ΐ 2 外 #: Ruler or combination of appearance dimensions may be selected. ^ The distance setting of the field with the smallest standard deviation, i: = r, 3 °, has the smallest variability ( (Or-caused; maximum) is the focal length center. If you can choose it, the parabolic curve can be equal to the data in Figure 2. The focal length center is determined by the minimum value of the parabola, that is, the slope of the curve is zero. In the example, you can

Page 33! 231358 V. Description of the invention (3〇_ Knowing that various statistics or distance points are used to obtain a more precise focal length center, which can be used in the intra-differential focus measurement technology. These methods are the existing y as shown in Figure 2. Those who are located on the center of the focal length of the micron may be exposed to the field 20, each field, in this example, the field is -0.1 (〇 · 〇) The domains are different. For the purpose of obtaining the position of the zero focus distance, etc., the engraving printer, rendering, program control, or similar values are related to the focal point of the method as provided by the present invention. This is the second preferred embodiment of the present invention; the focal length center corresponds to it. A selected parameter. Here is a fact: a theoretical information library or a diffractive structure (or a randomly selected fish ^ '有 从All changes or the degree of consistency in a certain field are determined = s) Comparison of obtained diffraction identification marks is determined by these diffraction identification marks :: The variability may not be directly measured by the judgment of the visual person =, however, This method requires the operation of any It is relatively slow to analyze the square and go to 3. Therefore, the variability of other marks. Such squares can be used to measure these diffraction identification methods such as the mean square error method (MSE) or brackets. However, it is not limited to, The statistical method of Euclidean distance measurement. This root error method (SE) 'and its weighted average, the sum of averages, and other large methods also include the characteristics of average arithmetic, plus difference. For example, in a :: = method to describe the difference between the diffraction identification marks, the greater the difference between its cm mean square? If the error is optional, the intensity of the identification mark is ^: ^ ^ Yue Sui is another, which can be used as a judgment of variability. ^ The range of values related to the center of focal length ^ can also be analyzed in a similar way

Page 34

1231358 V. Description of Invention (31). A series of diffractive structures, such as diffractive structures 3 0, 3 0, are preferably generated in a range of different focal lengths that have been determined and surround the center of the focal length, with the dose gradually changing from one structure to another in the form of a step Of it. The result is a series of diffractive structures each at a different but known dose. The two-column diffraction identification marks are then obtained for each diffractive structure, which is based on the radiation source as a basic tool. The series of diffractive identification marks produced by it can be analyzed as described above, for example, the difference analysis method is used to identify the diffractive identification. The generated focal length center can be plotted against the dose, so that the effect of the S-quantity on the focal length center can be obtained. In this way, the dose setting with the most sound focal line can be determined, so that a dose setting that has the least influence on the focal length curve or focal length-depth can be selected. , The whole, etc. of the present invention. In the center of the focal length, the center of these tools, the diffraction of the ray is limited to the use of tools. The angular resolution of wafers divided by angles will increase when the field is still there. Although there is still a technology, the radiation source can be scattered by the light and astigmatism of a specimen, and the aberration and astigmatism of the tool can be Sex. After that, however, there is little variability. It is clear to find a metrology builder who is looking for a focal length source, and how can it be done, including but not measured spectrum or wavelength-any tool can be consistent, such as the tilt of the stage, the variable field within the grid of the lens distributed on its wafer The variability of the grid will obviously show the most such overall non-uniformity of the grid in the field, which is effective in this case. It should be used to capture the radiation with t rays as the basis to intercept or penetrate a certain diffractive junction. In other words, any radiographic method can be used for this technique. Light reflection, ellipsoidal measurement, or polar-analytic segmentation tools and / or light energy can be used. More tools include

1231358 V. Description of the invention (32) Produces a response function such as a tool parameter or tool combination that causes a certain diffraction identification mark. Diffraction identification parameters suitable for these technologies include, but are not limited to, photoresist grids, and etched film stack grids. , And this technology. It can also be used to monitor drift in production settings and / or and / or stack thickness. When monitoring the diffractive structure 30, 30; ^ the selection parameter or the degree of change of the identification mark, if the calculated: f difference exceeds a certain value, the program may check the amount of drift.二 "Two means that a certain-process transfer should be investigated. In addition, as a kind of-general ^ order balance measurement method, measuring the structure across the entire wafer, and = = the difference between the identification targets also Can be used as a pair: r-'i two ways. A low degree of change in diffraction identification marks indicates program consistency, while a high degree of change in diffraction identification marks indicates a process such as semiconductor manufacturing; more = The steps and wafer types are exemplified here: the engraving printing, etching, and metal-cladding steps. The related steps require the use of a variety of different wavefronts. — First, remove the analysis of the focal length. It may have an adverse effect on the diffraction of the experiment; the filter of 3 ij is a measure of the goodness of the degree of matching based on the poor identification of the theoretical diffraction identification marks. Those who have a good match can be discarded from this analysis.

Mingwu Mingxi > I use the same as deciding that the first step will be to find a photoresist treatment process step. The square focal length of the present invention is the one of extreme importance for this step. However, the thin film stack: = T of the processing line, and the optimal focal length of the metal grid determines the /, or

Page 36 1231358 V. Description of the invention (33) '一 "' One — It is determined for the "optimal etching" and conditions of the etching process. The method and equipment of the present invention can also be used for quality control testing. For the purpose, including the analysis of the focal length center determined by other methods. This may be combined with an angle f-analytical light scattering meter to achieve 'the above-mentioned ones, including its related electric month "wide system" or other suitable devices. Make the described measurement method. If an angle-analysis light scatterometer is used for the periodic structure, the diffraction identification mark is divided into different diffraction orders of angular positions and described by the grid / equation: sin 〇i + sin θη = ηλ / d (3) where A is the angle of incidence, is a negative value, & is the angular position of the nth diffraction order, 'is the wavelength of the incident light, and d is the space period or the distance between the diffractive structures. Therefore, it can be seen that for the diffraction order of the zeroth order or reflection, the incident angle is equal to the angular position of the diffraction order of the reflection. However, diffraction orders other than the reflection diffraction order, or general light scattering or diffracting, may also be used, and the appropriate angular position is as proposed above. A similar system deals with other modes that produce diffraction identification marks, so that with any mode that produces diffraction identification marks, it is possible to reflect the diffraction order or some higher diffraction order or general light scattering or diffraction. Useful. For example, in column j, in a wavelength division device, the odor angle may be maintained: letting the wavelength be determined; and I changing, and solving the equation gives% at a given value of n.方法 The method and equipment of the present invention may also be used to determine the center of focus, whereby the center of focus is adjusted in any suitable manner, including the use of a computer control system 'and the method of the present invention is used to determine when an acceptable or optimal 隹

1231358 V. Description of Invention (34) " "-The distance has been decided. This adjustment may be achieved in a dose variation, or other manner known in the art. The invention can also be used for automatic or automated decision and adjustment of the focal length center; it utilizes an autofocus control system, where the input of at least one autofocus control system includes the variability of the selected consistency measure. Therefore, the present invention can use a program control which is a focal length. η-In the embodiment, the present invention provides a diffraction identification mark: and a controller. The controller computer can receive 彳 from the measurement device.佌, and can determine one or more parameters of a program step, such as focal length or dose, as a field variability of a large number of different parameter fields ... Determining its variability may include direct measurement; changing the size of the theoretical model of the identification mark, and the Jiajia adaptor theory diffractor computer may receive and be available at the same time. The brush control device, for example, controls a certain type of stereotype. The present invention may therefore be provided as the remaining number. Modification of certain-program feedback at a certain step; and control technology of feedback, in order to modify the theoretical or empirical setting limit three = for example, the field variability bit parameter may be based on the predicted parameter 外 2 outside, Within the limits of a certain procedure step. Since then, the program can fall in a certain moxibustion field where one of the program steps is desired. The internal variability is one or more fields that are modified theoretically or empirically. And decide, and can be located in the field or concerned ^, compare. The internal variability of the field may be the variability of a certain subcollector. All or

1231358 V. Description of the invention (35) This diffraction is used, so that the processing steps need more processing procedures, or a series of steps to determine if the downstream processing processing steps are already implied that a certain change is not a measurement method In order to use the variability of the circle processing process, the mark amount is divided separately. Generally speaking, in this step, this procedure changes the key to the midfield of the procedure. Directly and physically monitors the subsequent processing steps. Potentially other steps are used in the development of the sequence. <Dunguojia refers to this guide but still @ 加 定 —The step measuring device can use a method that can be used to limit the scope of the party. For example, after the exposure, the increase in the variability of the focal length range that is obvious at any wafer processing site has a harmful result. It is similar to the focal distance setting, which is related to the target degradation. Therefore, this method can be used to add images to the wafer, after wafer processing, after baking; In such a way that this variable. Although this provides a field inside the entire crystal, Example 1

According to the present invention, there are five fields exposed to different micron increments from -0.2 micron to +0.2 micron. For each field, a 5x5 array of diffractive structures is similar to that described in Figure 1B / Figure 3, and the key dimensions (c β) of the diffraction structure in each field are shown graphically. Figure 4 shows the 1-σ (Sigma-Sigma) 4 quasi-bias values of the key dimensions of all 25 diffractive structures in each field. As mentioned above, the optimal focal length is achieved by setting a focal length of -0.1 micron. Regarding real-time focus monitoring and program control, it is important to be able to change it before the focus / indication shift becomes too large. It is therefore advantageous to monitor the

1231358 V. Description of the invention (36) The key dimension of a subset of isodiffractive structures or its?: Isodiffractive structure shows a change in its selected appearance scale: ::? = The larger of all 25 diffractive structures. For example:; = 二 = The change of the size to the focal length: It may be that the person who is located in the center of the 30% domain is much larger. Figure 5 simultaneously plots the structures of the projections (ie, the same data as in Figure 4) because the J ^ column has only a 1-σ standard deviation in the fifth row. Xi ..., the slope of the set of 5 structures is larger than that for 25 structures: 隹 ^ Minimum value is easier to determine. In addition, since only this one diffractive structure is monitored, if compared with all 25 structures, its focal length Γ ^ will be detected more quickly due to its steeper curve. This method Z is also used to more accurately determine the focal length center of a case. The case is right and right; the change curve of the focal length derived from all diffractive structures in each field is close to flat, without a clearly obvious minimum value. p, Figure 6 is a graph of the same data as shown in Figure 4, but instead of plotting the standard deviations of 25 ^ values in each f domain / pound, the plots of the 25 CD values are plotted ( (Maximum value minus minimum value in nanometers). It can be seen that both sets of data provide the same result, that is, its focal center appears at a focal length of 10 μm. Example 2 According to the second preferred embodiment discussed earlier, FIG. 7 depicts diffraction identification marks generated by five spaced-apart diffraction structures in three fields taken at different focal length values. In this example, it can be easily checked by inspection. Page 40 1231358 V. Description of the invention (37) It is judged that the degree of focalization at 0 · 0 micron is the smallest. The range of the variation of the diffraction identification mark in the field in Figure 7 is shown in Figure 2 / icons, showing the strength of each field identification mark. This method can be used instead of the above-mentioned field, that is, the field is located at the optimal focal length. Those program control and focal length drift monitoring J! What are the consistent examples, including example 3? A typical mask is engraved and changes in size over time. The range of 3-σ (si gma) and the focal distance shot structure are the same for each cover and procedure. However, this leads to a higher level of awareness that if the control limit of acceptable nanometers exceeds the control limit, it may be used in the 13th and 21st. Program control application examples include when multiple wafers are treated the same ~ Monitor the selected external dimensions during manufacturing using the same procedure. Figure 8 depicts the selected degree measured during the hour (in this case, the change in c D is measured in degrees, preferably including the illumination field previously determined in accordance with the method of the present invention). Because the same wafer is measured in the same field, and each wafer is printed with the same mask, the variability of each wafer should be related to the focal length of the printing tool may drift with time. The dimensional variability is shown in FIG. 8. According to previous knowledge, the change of private sequence control may be 2.4 nanometers. Therefore, 2 4 degrees is shown in FIG. 8 as a dotted line. The degree of change is usually very small 'if any. If the degree of change is already considered a problem, a certain focal length is changed ^. Within the data for this example. Program correction follows data measurement for data hours. In both cases A, the CD is variable

Page Ή .1231358

K is reduced from the next measurement size determiner 'to the familiar control procedure. When the change data exceeds the control limit for the first time, there is no order: ± is being used. This is to avoid over-correction for possible non-existent problems, as it is purely a random variation outside the control limits. The exact rules of control vary from program to program, but the concept of controlling focal length by using only a single domain is the same. A similar example can be derived for the application of the method of the present invention. Its focal length is determined by comparing the degree of change from all diffractive structures contained in the field, or a subset of them. In cases where the degree of change exceeds a certain amount of change in a predetermined identification mark (for example, expressed in units of mean square error), a corrective action is best used. The method of this example may be used as feedback and feedback control technology to modify a certain parameter of a program step. For example, 'Variability within the field exceeds the limit set theoretically or empirically, a parameter of a certain procedure step may be modified based on speculative parameters, so that the procedure falls back to the desire Inside the restricted range. This procedure can be automated, and it is possible to use a computer-controlled system that will collect diffraction identification marks, perform identification changes on the identification marks of the diffractive structures in the field, and, based on the change analysis, print the engraved plate One or more of the desired parameters in the device exhibit a certain change, for example, the dose is focal length. Although the present invention has been described in detail with respect to these preferred embodiments, other embodiments may achieve the same results. Variations and modifications of the present invention are bland for those skilled in the art ’, and all these will be different

Page 1212358

Page 12 1231358

"The illustration of the special companion k is included in the specification and becomes one of them", which describes one or more embodiments of the present invention with a drawing, and is used for explaining the principle of the present invention together with its text. The illustrations are for description only

The purpose of inventing one or more preferred embodiments is not to be construed as limiting the description in the illustration: X FIG. 1 A 4 ¾, ', said a wafer with some fields and light at different focal lengths; Figure 1 B traces one of the fields in Figure 1 A and includes some diffractive structures; Figure 2 is the standard deviation curve of the critical dimension (CD) for each field diffractive structure; Figure 3 A-3E series Graphically depict the degree of change of CD values in different fields according to Example 1 of the present invention; FIG. 4 is a standard deviation curve of all CD measurement values according to each field in Example 1; FIG. 5 is a graph of FIG. 4 , Plus the standard deviation curve of the more variable measured CD value subsets in each field according to the example 丨 Figure 6 is the graph of the change amplitude of all measured c D values in each field according to example 1 ; FIG. 7 is a graph illustrating comparison results of diffraction identification marks based on fields of different focal lengths in Example 2; and FIG. 8 is a graph illustrating measurement values of program control quantities according to Example 3

Page 44

Claims (1)

1231358 1. A parameter for measuring and engraving printing equipment includes the following steps. The method includes a substrate containing a plurality of field fields, each field distance value being exposed, and a plurality of number diffraction structures. The engraving printing equipment is formed on the substrate at different focal lengths; the x-printing program uses a radiation source-based tool to measure its diffraction identification marks in a number of diffractive structures; '&quot; Determine the variability of each field by the number of diffractive identification marks located in that field; and the variability associated with those fields compared in the construction to determine a desired parameter . ~ The engraving printing device 3 · The method as described in item 2 of the Shen Jing patent scope, wherein the diffractive structure includes a grating. 4. The method according to item 1 of the scope of patent application, wherein the substrate comprises a semiconductor wafer. 5. The method as described in item 1 of the scope of patent application, wherein the radiation source is a basic tool including a light source as a basic tool. 6. The method according to item 1 of the scope of patent application, wherein the light source-based tool includes an incident laser beam source, an optical system focuses and scans the laser beam through a range of incident angles, and a The detector detects the diffraction identification mark formed for the entire measurement angle. Page 45 1231358 VI. Scope of Patent Application 7 The method described in item 6 of the scope of patent application for molasses, wherein the tool of the light source β '' includes a scatterometer that analyzes by angle-division. The method described in item 5 of the scope of patent application, wherein the * line light source is a basic tool including a number of laser 9 rabbits. The method described in item 5 of the scope of patent application, wherein the light source is the tool of the present It includes an incident wide-spectrum light source, an optical system that focuses light and emits light at a range of incident wavelengths, and a detector to detect diffraction identification marks formed for the entire measurement wavelength. 10. The method as described in item 5 of the scope of patent application, wherein the light source is a basic tool including an incident light source, whose constituent parts are those that change the amplitude and phase of the polarization, and an optical system focuses the light And to identify the phase of the sign. The radon and radon detectors use the method described in item i of the scope of the patent application to detect the formed windings. The identification mark includes a wide-band radiation source as a basic tool for phase in-line measurement. It operates at a certain fixed angle and can be changed to a variable scanning angle φ. The angle is $ 8, and the angle of incidence θ or some a, as in the scope of patent application! The method described in item 1, wherein measuring a diffraction identification mark includes a single-wavelength radiation source as the basic tool and position measurement, which operates at a fixed angle and a variable scan. The angle φ. The angle of incidence ㊀ or 13 is the method described in item 丨 of the patent application scope, wherein the measurement identification mark includes a phase-based measurement of a separate multi-wavelength radiation source. 口 31358 ιτΓίί 利 范围 I 4. The method as described in item 1 of the patent scope of the application, wherein the diffraction identification log is a reflective diffraction identification mark. 15. The method as described in item 1 of the scope of patent application, wherein the diffraction identification mark is a conductive diffraction identification mark. 16. The method according to item 1 of the scope of patent application, wherein the diffraction identification mark is a reflection identification diffraction mark. 17. The method according to item 1 of the scope of patent application, wherein the diffraction identification mark is a higher-order diffraction identification mark. 18. The method as described in item 1 of the scope of the patent application, wherein the diffraction identification mark is a measurement value of general light scattering or diffraction. 19. The method according to item 1 of the scope of patent application, wherein the desired parameter is the focal length center. 20. The method according to item 1 of the scope of patent application, wherein the desired parameter is a dose. 21 · The method as described in item 1 of the scope of patent application, wherein the value of the desired parameter of the engraving printing device is determined by the value of the desired parameter associated with the field having the minimum feasibility of the diffraction identification mark. . 22. The method as described in item 丨 of the patent application range, wherein the determining step includes measuring the intensity range of the diffraction identification mark for each field, and the identification mark is a diffraction measured by a plurality of numbers located in the field The identification mark was obtained. -23. The method as described in item 丨 of the patent application scope, wherein the determining step includes calculating a statistical measure of the variability. 24. The method as described in item 23 of the scope of patent application, wherein the statistics I I231358 VI. The number of patent application scopes is the diffraction recognition 2 5 · Different Jiaozhu shot structures as known in the application, and 2 6 · If the application shot structure contains the same set as the known 27. If you apply for special The method described in item 1 of the root mean square error margin of the sign, the method further includes forming a small number at the distance setting and known different dose settings and determining the effect of its dose on the focal length. / The method described in item 25, wherein the number of diffractive structure sets of the same number of different focal lengths is the same as the different dose settings of the diffractive structure set. The method described in item 1, wherein the step of determining includes providing a theoretical diffraction identification information database generated by a theoretical diffractive structure, and determining the diffraction identification mark for each measurement in the information database. A best-fit theoretical diffraction identification mark; associating a selected appearance dimension of the best-fit theoretical diffraction identification mark with the measured diffraction identification mark; and determining its relationship with each field The variability of selected external dimensions associated with most diffractive structures inside the field. 28. The method as described in item 27 of the scope of patent application, wherein the selected appearance dimension is a critical dimension (CD). 29. The method according to item 27 of the scope of patent application, wherein the selected appearance size is a cross-sectional area. 30. The method described in item 27 of the scope of patent application, wherein the selected appearance size is a cross-sectional volume. 31. The method according to item 27 of the scope of patent application, wherein the selected appearance 1231358 VI. Scope of patent application Dimensions are the product of two or more external dimensions of theoretical diffraction structures that match the theoretical diffraction identification standard. 32. The method as described in item 27 of the scope of patent application, wherein the determining step includes measuring, for each field, a range of selected appearance dimensions associated with a diffractive structure measured with a power majority located in the field. 3 3. The method as described in item 27 of the scope of patent application, wherein the step of determining includes calculating a statistic measurement of its variability. % 34 · The method described in item 33 of the scope of patent application ', wherein the measurement value of the statistical value is the standard deviation of the selected appearance size ° 3 5 · The method described in item 1 of the scope of patent application Where the diffractive structure includes a latent image diffractive structure. / 3 6 · If the scope of patent application is the first! The method of clause 2, wherein the substrate has not been subjected to a developing process. 3 7 · —Program control method for a focal length center in a printing plate 'This method includes the following steps: The focal length center of the printing plate is determined according to the method described in item 9 of the patent application scope; and The focal length setting of the printing equipment is adjusted to the determined focal length. 038. The method described in item 37 of the scope of patent application, wherein the adjustment step uses a computer-based control system. 39. The method according to item 37 of the scope of patent application, wherein the adjustment step includes an autofocus control system, at least one of which has an autofocus control Page 49 1231358 Scope of Patent Application System Rouwei _ 4〇. Assumptions include a measure of the amount of minimum variability. The clause recognizes the method described in item 37 of the patent scope, wherein the adjusting step 3 is performed at the same time as the time, and is measured by a number of diffractive structures located in the selected field. Various variability. After θ * The method described in item 40 of Shen Zhiqing's patent scope ’, wherein the selected field is not previously determined to be located at the center of the focal length. 4 ^ * Female Master Shi Kui, the method described in item 40 of the profit-seeking range, in which if the variability exceeds the control limit of a certain blind spring +, top priority, the focal length of the printing equipment is Broken adjustment. 4 3 • A method for program control in a engraving printing device, the method includes the following steps: exposing a 4 &quot; Xu diffractive structure with the engraving printing device in a certain field, exposing on a series of wafers; using a A radiation source is a basic tool for measuring the diffraction identification marks of various diffractive structures on a large number of 4 series wafers in a certain field; determining for each wafer that it can be obtained by a number of diffractive structures. Denaturation; and, ^ Compare the denaturation associated with these wafers to control one of the desired parameters of the engraved printing. 44. The method as described in item 43 of the scope of patent application, further comprising the step of adjusting at least one desired parameter of the stencil printing device in response to the comparative variability associated with the wafer. 45. The method as described in item 44 of the scope of the patent application, wherein the adjusting step includes a comparison with an empirically determined threshold for degeneration. 1231358 6. Scope of patent application 46. The method described in item 44 of the scope of patent application, wherein the adjustment step includes a comparison with the variability limit determined in a theoretical manner. 47. The method as described in item 44 of the scope of patent application, wherein the at least one desired parameter comprises a focal length or a dose. 48. The method described in item 43 of the scope of patent application, wherein the diffractive structures are monocyclic, bi-periodic, multi-periodic, or non-periodic structures. 4 9. The method as described in claim 48, wherein the diffractive structures include a grid. 50. The method as described in item 43 of the scope of patent application, wherein the wafers include semiconductor wafers. 51. The method as described in item 43 of the scope of patent application, wherein the radiation source is a basic tool including a light source as a basic tool. 52. The method according to item 51 of the scope of patent application, wherein the light source-based tool includes an incident laser beam source, an optical system focuses and scans the laser beam through a range of incident angles, and a The detector detects the diffraction identification mark formed for the entire measurement angle. 5 3. The method according to item 52 of the scope of patent application, wherein the light source-based tool includes a scatterometer based on angle-division analysis. 54. The method as described in item 51 of the scope of patent application, wherein the light source-based tool includes a plurality of digital laser beam sources. 55. The method described in claim 51 of the scope of patent application, wherein the light source-based tool includes an incident wide-spectrum light source, an optical system that focuses light and emits light at a range of incident wavelengths, and a detector Page 51, 1231358 Application for patent scope 1 Detection of diffraction identification formed for the entire measurement wavelength. As described in the method of patent application No. 51, ^: ° The basic tool includes-a kind of incident light source, : Light source, the amplitude and phase of a P polarization, an optical system that focuses light and emits light at a range of incident phases, and a detector to detect the phase identified by diffraction and identification. 57. The method according to item 43 of the scope of patent application, wherein measuring a diffraction identification mark includes phase measurement based on a wide-spectrum radiation source, and operates at a fixed angle, a certain Variable incident angle Θ or some Variable scanning angle Φ. 58. The method according to item 43 of the scope of patent application, wherein measuring a diffraction identification mark includes phase measurement based on a single-wavelength radiation source as a basic tool, and operating at a fixed angle, a certain A variable incident angle Θ or a certain ^ variable stroke angle Φ. 5 9 · The method described in item 43 of the scope of patent application ', wherein measuring a diffraction identification mark includes phase measurement using a separate multi-wavelength radiation source as a basic tool. 60. The method as described in item 43 of the scope of patent application, wherein the diffraction identification flag is a reflective diffraction identification mark. 6 1 · The method as described in item 43 of the scope of patent application, wherein the diffraction identification 榡% is a conductive diffraction identification mark. 6 2 · The method described in item 43 of the scope of the patent application, wherein the diffraction identification flag is a diffraction recognition mark of the reflection order, 63. The method described in item 43 of the scope of patent application, Where the diffraction recognition 1231358 VI. Scope of patent application The mark is a higher-order diffraction identification mark. 64. The method as described in item 43 of the scope of the patent application, wherein the diffraction identification mark is a certain measurement value of general light scattering or diffraction. 6 5. The method described in item 43 of the scope of patent application, wherein the determining step includes measuring the intensity range of the diffraction identification mark for each wafer, and the identification mark is located on the wafer in the field by a large number of numbers. Obtained from the measured diffraction identification mark. 6 6. The method as described in item 43 of the scope of patent application, wherein the decision step includes calculating a statistic measurement of the variability. 6 7. The method according to item 66 of the scope of patent application, wherein the statistical measurement value is the root mean square (r ο 〇 t m e an n s q u a r e) error of the diffraction identification mark. 68. The method described in item 43 of the scope of patent application, wherein the determining step includes providing a theoretical diffraction identification target information database generated by a theoretical diffractive structure; and the measured diffractions in the information database Identification of the theoretical diffraction identification mark that determines an optimal fit; ", a certain selected appearance size is linked; and a large number of diffraction changes on the wafer in the field. Method, wherein the selected appearance will be the best The fitting theory diffraction identification is related to the measured diffraction identification mark. It is possible for each wafer to determine the selected appearance size associated with the structure. 6 9 · As described in item 68 of the scope of patent application, the size is Key dimensions (CD). 1231358 6. Scope of patent application 70. The method described in item 68 of the scope of patent application, wherein the selected appearance size is a cross-sectional area. 71. The method as described in item 68 of the scope of patent application, wherein the selected appearance dimension is a cross-sectional volume. 72. The method as described in item 68 of the scope of patent application, wherein the selected appearance size is a product of two or more appearance sizes of a theoretical diffractive structure adapted to a theoretical diffraction identification mark. 73. The method as described in item 68 of the scope of patent application, wherein the determining step includes measuring, for each wafer, a range of selected appearance dimensions associated with a plurality of digital diffraction structures on the wafer located in the field. 74. The method as described in item 68 of the scope of the patent application, wherein the determining step includes calculating a statistical measurement of the variability. 75. The method as described in item 74 of the scope of patent application, wherein the statistical measurement value is the standard deviation of the selected appearance dimension. 76. The method as described in item 43 of the scope of patent application, wherein The diffractive structure contains a latent image diffractive structure. 77. The method as described in item 43 of the scope of patent application, wherein the wafer has not been subjected to a development process. Page 54