TW201209371A - Method and system for optical metrology optimization using ray tracing - Google Patents

Abstract

Provided is a method and system for determining a profile of a structure using an optical metrology system that includes an optical metrology tool, an optical metrology model, and a profile extraction algorithm. The method comprises selecting a number of rays, selecting beam propagation parameters, determining beam propagation parameters, calculating total intensity and polarization of the diffraction beam, calculating a metrology output signal, and extracting profile parameters. Also provided is a method for determining profile parameters of the structure optimized to achieve accuracy targets, the method comprising: setting accuracy targets; selecting a number of rays and beam propagation parameters, measuring a diffraction signal, generating a metrology output signal, determining an adjusted metrology output signal, concurrently optimizing the optical metrology tool model and the profile model using the adjusted metrology output signal and a parameter extraction algorithm.

Description

201209371, invention description [Technical field to which the invention pertains] ... is optimized by the household structure. Nine bundles of the parameters, while considering the knot [Prior Art] Optical measurement system is to direct the incident beam to the diffraction signal formed on the workpiece, and analyze the measured ρ, 、, 冓, 测 所. The work piece can be a tiger's to determine the various features of the structure, usually m book = warranty = or magnetic media. In the case of the fabrication of the semiconductor wafer, the operation of the semiconductor wafer, L: into two, : / /, crk collection. The receiver analyzes this diffraction, ^ ^ fine pairs. In the database, every time the charm is successful, the diffraction signal _ measurement number and data ====== can be 敎 with the analog diffraction signal period grating_. Slim to produce with accurate table _ _ ^ ^ = Xie turn to the same __, should the cost is near low? Further, the size of the spot is reduced by ', and the size of the spot is reduced.' When the lithography node is 3 〇 nm or lower and the 201209371 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Furthermore, in the modeled optical measurement tool, the accuracy of the wheel corridor parameters of the IfίIf ϊ ' 有关 有关 physics physics, ϊ 其 „ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ The method of the profile is dead, the line is determined, and the beam travel parameters are determined; ^, each light in the selected light of the detector is determined to be the first _ intensity polarization of the diffracted beam on the beam of the beam; The sample knot "the light in the selected light - the detection - the detection signal = the output signal, the calibration data, the total intensity and polarization of the ray beam of the rim; (e5) the letter from the total intensity meter Test, enter the sign j«丨, 彳 output by the tiger, (c6) profit [implementation] round or substrate to show sales, process crystal 201209371 V7 uU L唬57 transfer to the processor 53. processor 53 take The simulator 6〇 measuring the junction f shows the knot, assuming that the “difference” and “two-f” different Maxwe11 equations are used for the immediate diffracted signal generator (in the regression example, the simulation 11(8) is selected to produce the best match. Fence + inch =: the actual cross-sectional shape of the feature portion corresponding to the sample structure 59 and 40 may use a reflectometer, ellipsometer meter, or other optical device, wire or weaving the old file worse optical measurement charge ... 'This document is incorporated herein by reference: U.S. Patent No. hair 69439〇〇

The name of the month is "GENERATION OF A LIBRARY OF PERIODIC GRATING DIFFRACTION SIGNAL", and the award date is September 13th. . The Maxwell equation is applied and the Maxwdl equation is calculated using numerical analysis to produce an analog diffracted signal. It should be noted that various numerical analytical methods are available, including various RCWAs. A detailed description of the RCWA can be found in the following document, which is incorporated herein by reference: U.S. Patent No. 6,916,626, entitled "Caching of intra-layer calculations for rapid rigorous coupled -wave analyses"' application date is January 25, 2001 Day, the certificate is May 10, 2005 曰 0

The analog diffracted signal can also be generated using a machine learning system (MLS). The MLS is trained with known input and output data before the simulated diffracted signal is generated. In an exemplary embodiment, the 'generating diffracted signal generation' can be performed by MLS 201209371 using a machine learning algorithm such as backpropagation, radial basis functions, support vectors, kernel return, and the like. For a detailed description of machine learning systems and algorithms, reference is made to the following documents, which are incorporated herein by reference: U.S. Patent Application Serial No. 1/6/83, entitled "optical metrology of stmctures f〇 Rmed 〇n semic〇nduct〇r wafers using machine learning systems”, applied for June 27, 2003. 2 is an exemplary block diagram of an optical metrology system in accordance with an embodiment of the present invention. At least the optical optical measurement may be included, and at least two optical outputs from the secondary light system may be transmitted to the illuminator subsystem (10). From = = the learned output (1) can be propagated to the selector sub-system k. The human system 115 can transmit at least two signals 116 to the beam-generating light. The at least two reference outputs 126 can be supported by the platform base (10). Circle 1G1 "Up; Sensor 104. You 4 卞 岫 晶 晶 晶 晶 Ϊ Ϊ 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 At least f "low A 〇 I" (A0I is the angle of incidence) under the worry of 131, when in the second mode "high A 〇 I two! Π 1 control on the brother one path 132. When the first optional The reflective sub-system operates at a lower level when the first mode "lower mode is at least lower than the second path from the beam generator subsystem 120". • the output 141 can point to a high angle convergence, at least a second of the human system.肷丄J from the first beam generation ^ owing system 12 < the output of the death points to the first reflection subsystem 140; from the first - path (3) an output 141 can point to at least the high angle Poly-June, the at least the human system 1 as the second way secondary system 12 (sub system 135. Or, in addition to "low 曰 曰 low angle convergence Other modes and other configurations are used. In addition to the mode of "~ back to A01", the equivalent measurement system 100 outputs at least two times of the first tenth focus sub-system 145. 1 from the high angle 201209371 At least φ 1 κ ι 统 135 135 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , φ φ φ φ /@糸, first 150, first optional reflection sub-system 160. recognize s丨, - 糸, first 100 to the first mode "low Α (31" operation, from im round im to ^, an output 156 It can be pointed to a high angle to receive the female day angle of incidence. In addition, the high center system 155. For example, a high angle output 156 can be used, and the high degree system 155 can process the secondary system from the wafer 1〇1. 150, = = can provide the output 151 to the second reflective reflection secondary (10): the output of the second optional system (9) 1521 is allowed to be from the second reflective sub-system 詈 (four) aunt, zhe The phase loss is passed through the second optional reflection subsystem 160. The output 166 can be operated as a high A〇1 operation from the wafer 101 ϋ low angle collection subsystem 165 °, for example, can be used when the lower 165 ^ = rectification, from the second optional reflection Ge 糸 棋 OB AOI sub system no. Di ray - human system 160 output (6) can point to the analyzer ^ equivalent When the system 100 operates in the first mode "low A 〇 I", when the high, the (four) wafers are 〇 1, the low ^ corner ^, the analysis can be analyzed by the _ system i7G, the system can include At least two measurement systems (7). f 175 may include at least two detectors operating in the ultraviolet to visible region of the body. The at least two illumination systems and the imaging subsystem 182β may also include at least two imaging subsystems 182. At least two

8 S 201209371 * System 184. (Description output ι86) In some embodiments, the measurement system 100 can include at least two autofocus subsystems 190. Alternatively, other focusing techniques can be used. At least two secondary systems (105, no, 115, 12〇, 125, 13〇, 135, 14〇, 145, 150, 155_, 160, 165, 170, 175, 180, may be used when performing the measurement of the structure) At least two controllers (not shown) of 182, 195). The controller can receive actual signal data to update the secondary system, processing components, processes, recipes, rims, images, patterns, and/or model data. At least two secondary systems (105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 182, 190) may use at least two semiconductor device communication standards ( SECS) messages to exchange data, read and/or remove information, feed forward and/or feedback, and/or transmit information as SECS messages. Those skilled in the art will appreciate that at least secondary systems (1〇5, 110, 115, 12〇, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 182, 195) may include the required computer and memory components (not shown). For example, a memory component (not shown) can be used to store information and instructions for execution by a computer (not shown), and can be used to store temporary variables when the various computers/processors in the measurement system 1 execute instructions. Or other intermediate information. At least a secondary system (105, 11 〇, 115, 12 〇, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 182, 190, 195) may be included to A device for reading data and/or instructions on a computer readable medium, and may include means for writing data and/or instructions to a computer readable medium. In response to the computer/processor executing at least two of the at least two instructions stored in the memory and/or received in the message, the measurement system 100 can perform some or all of the processing steps of the present invention. These instructions can be received from other computers, computer readable media or network connections. Furthermore, at least the secondary system (105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 182, 190, 195) may comprise a control application, Graphical user interface (GUI) component and/or database component. It should be noted that the 'equivalent measurement system 100 operates in the first mode "low A0I" with high angle of incidence data from the wafer 101 to the measurement subsystem 175 k. (outputs 166, 16 162, 171) And the Equivalent Measurement System 1〇0 operates in the first mode “High A〇I” with low incident angle data from the wafer 101 to the measurement sub-system 175 for the entire length of the path. 201209371 = Output 156, m ' 152, 162 171) 'The beam is called (one or more) diffraction signals. 0 Figure 3 shows the architecture of the optical quantity of the prior art, which is commonly used in optical measurement tools (such as photometric devices) and processing. Between modules. The optical measuring device includes a light source for generating a wire and directing the light onto the structure, and for detecting the light radiated from the structure and converting the side light into the measured diffraction signal. . The processing module 312 receives the measured diffracted signals from the optical metrology tool 3〇4, but in particular receives the measured diffracted signals from the encoder to analyze the structure, such as the rim of the structure. The splicing luminosity measuring device, which utilizes various signal parameters, can provide the measured ray" 唬. General; 丨 308 is used to utilize a group of standard signal parameters, and the measured signal processing module 312 is provided. The standard signal number of the set includes the inverse reflectance (Oreflectance) parameter, which characterizes the change in light intensity when the light is reflected on the structure, and the change in the amplitude of the polarization when the light is reflected on the structure. When the optical metrology tool 304 field measures the reflectance of the change in light intensity (e.g., a spectroscopic reflectometer), the measured diffraction signal is supplied to the processing module 312 by the 'reverse number of the interface 3G8 fine signal parameter fresh group towel'. When the optical quantity measurement is used to measure the change of the light intensity and the change of the light polarization phase, the number of reflections of the elliptical rotation is the number of the money, and the reflection of the sharp signal of the sharp signal is transferred to the polar polarization parameter. Including: the first value of the _, == Ϊ, the absolute value of the squared S-polarization and the difference between the Ρ-polarized light, U(s), characterized by two normalized to the s of R The imaginary part of the complex reflection robust interference; and the third parameter /,, = ί=包之Number = s 201209371 Magic is often used to characterize the polarization state and light intensity of light in an optical instrument. Or 'Using the J0nes matrix equation, the same purpose can be achieved with a homology matrix. Through the following relationship, the Stokes parameter and the homology matrix Related: (s〇s, s2 s3)=(Jxx+Jyy j^+jyx KJyx.Jiy)) (1) For a simple round polarimeter, the elliptical polarization parameter p = tanv^△ is usually used. In this case, Stoke The relationship between the S parameter S & phantom pair to the common elliptical polarization parameter p = tanv/y can be expressed as follows: (^0 S2 53) = /0i?(l -cos2i^ sin cos Δ sin2i//sinA) (勹, where : P = Ίεαχψ ei& = i = (3) K Es0 The ellipsic polarization parameter (P = tan"'A) can be characterized by three parameters to characterize the repetitive effect. In the simplest case, if there is no depolarization, This relationship can be expressed as (Μ SC) — (cos 2ψ sin 2ψ sin Δ sin 2ψ cos Δ) , } (4) χ = β =\ , ' (5) Photometric measurement used in optical measurement of semiconductor structures Devices that typically use a focused beam to produce small spot sizes (in the range of μιη). Therefore, for use A photometric measuring device for focusing a beam, the measured diffracted signal being an integral of the measured diffracted signal corresponding to all linear rays in the effective numerical aperture (ΝΑ) of the photometric device. Each of the rays in the NA corresponds to a specific incidence. Angle (A〇I) and wavelength. In addition, the absolute value of the complex reflection coefficient (CRC) squared & Γρ, and further parameters (R, Nsc), is a function of the angle of incidence (A0I), where R is defined as follows Reflectivity. Since the focused beam depends on the AOI, the focused beam is depolarized. According to this, the ellipsometric parameter (p = tan", A) is no longer sufficient to describe the characteristics of the focused beam. In addition, the definition of _(5) must be reconsidered, and should be 11 201209371 征化: to να borrowing ^^^^^^quot;2 is no longer equal to 1. Furthermore, 'depolarization is not limited to skew = limit'. It can also be the result of limited spectral resolution or other effects. :’’: an exemplary photometric device, the measured diffraction signal can be as follows:

1 = PSD-M-PSG (6) $ /5 ^ Ning PSD represents the polarization state detector for the polarized light Stokes parameter to i, PSG for the Stokes parameter generated by the polarization generator I. 1, Μ is the Muier matrix. Vector PSD and psG are not functions of A〇I and ^. For a given ray (with a given AOI and wavelength), the matrix can be written ^(ΑΟΙ,λ) 'Rp + Rs Rp~Rs Rp~Rs Rp + Rs 0 0 0

Re(Rsp) Im(Rsp) -Im(Jisp) Re(Jisp) (7)

Among them, P=k, J2 X, and L% are complex reflections. For a photometric device using a focused beam, the measured diffracted signal is the intensity integral of all linear rays and the center of the photometric balance and the sideband width near the wavelength. This integral can be calculated only for the Mueller matrix as follows: · / = [/(AOI, X)dQA0I άλ =PSD. [\M{AOI, X)dQA〇l ^)- PSG

Then, the generalization parameter (R, NSC) can be defined as follows: j(Rp + Rs)dQA0J dX

R N:

S

C \\da ^{Rp-Rs)d€lAOIdX ^{Rp + Rs)dQA〇I dX jlm(Rps)dnA01 dX J(Rp + Rs)dQA0J dX άλ j(Rp + Rs)dCi

dX

dX (8) (9) (10) (11) (12) Execute the above equation (8) around the center wavelength of the photometric device

S 12 201209371 Partial center wavelength data available, or at wavelength. ▲ The interface and signal processing module can convert and process the measured = spectrum. After the 'beauty is available' # 征化 indicates the average of the Π Π 上 上 上 上 上 上 上 上 上 上 上 上 上 上 上 上 上 平均值 平均值 平均值 平均值 平均值 平均值 平均值 平均值The _ characterization represents the half value of the normalized difference between the structure and the reflectance. s is the imaginary component of normalized positive iffίtr interference, ie not _bit component. c is the real component of the interference of the complex reflection coefficient of the ice gauge i, that is, the in-phase component. In addition 'make = linear, light time' and therefore the Stokes parameter two two miles (four) this 'Ns^ sr using equation (9) up, normalized Mueller matrix 1 -# 〇〇) 7 m^=-n 1 ° 0 0 0 C 5 ,0 0 -sc/ (13) Therefore, the measured diffracted signal can be characterized as:

I = PSD-(RM')-PSG _ ___ (14) . Normally 5, + and (R, NSC) are four independent parameters. The parameter (R, NSC) can fully describe the isotropic addition of a film. The reflection feature of the structure is not described but does not describe the polarization interactions that exist in the unequal (_〇tr〇pic) structure of the periodic grating. However, use 90. The azimuth eliminates the intervention of the polarization interaction effect. The parameter R $, M' or equal (R ′ Nsc) is a function of the measured state of the center A 〇 I, center wavelength, effective NA, spectral resolution. From the parameter ruler to μ, or the equivalent definition of (R, NSC), as long as there is information about the central enthalpy, the center wavelength, the effective enthalpy, and the spectral resolution available, these magnitudes can be simulated. The fit can be achieved using a database or technique of returning or machine learning in the program. For a detailed description of the interface of the optical measurement system, see the following document, which is incorporated herein by reference: U.S. Patent No. 7,048 229, entitled "rGENERIC mTERFACE F〇R ^ a- 13 201209371 OPTICAL METROLOGY SYSTEM" It is June 2, 2006. The upper structure size becomes smaller, and the factors that do not materially affect the accuracy of the measurement previously have an effect on the production. Furthermore, the assumption of 'turning the optical system to defend the wealth is no longer ## I θ 4 to 11Β is an exemplary architecture diagram, which is used in the optical measurement tool model for the age of the wheel 四 (4). The light used is missing _ illumination, detailed spectroscopic influence of the diffracted beam, beam side parameters. ^ m ^ Wei Wei's silk measurement tool architecture. I, only three rays, but the amount of light used for ray tracing can be a single ΐί 2 2ί more light. Optical metrology is shown using three light rays from the if optical measurement elements 426, 430 that are directed toward the workpiece 454. The workpiece 452 is placed on the motion control system 444, and the motion control 444 is used to adjust the illumination beam to focus on the workpiece 452. In this figure, the light passes through the optical tool _ illuminating portion 49G and the detecting portion ____ element, reaching and including the detector 484. The first illumination ray 408 /spoon,; 'transmits through the optical element 426 with the light 422, the optical element 426 ί4 Γ Γ Γ ,, or the fill patch ′ produces the output light 424. Output ray 454, ", element 430' produces output ray 430 at incident angle % at structure 426 light, line 416 "light 42", through illumination optics polarized min as described above, illumination optics 426 may include The output light, the line 432 is propagated to the focusing element (4), the second (four) is generated, and the light ray is on the structure 454. The cut surface 420 of the illuminating beam is ' and only f is traced through the system A representative light ray. Each light ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ #output illuminating light 436 to be combined with the incident angle set i and the element:: corner ΐ ray 458 ί, the measured light 458 propagates to the receive green ray 464, passes through the collecting optics 460 to become the % shot line 464' Output light 'line 47〇, and travel to detect light to collect

14 S 201209371 L484i iiif 466 may include a collimator, a fill patch, and/or a collection bias. The portion 494 may include other optical components, such as θι, which may include one or more detections. ::r - 456, · ί_= light = cow = degree, become a side ray. The side ray-, 士" first gives the 460, generates the detection light 462, passes through the collection and learns the second, and collects the 光线f ray 476 to travel to the collector 484. The film, also known as ^乍八^减_7 includes a collimator, a fill light, and/or a collection of polarized light to collect light rays; and a second detection beam of the cut surface 472 is a Huduo & or Xiugu detector.

The phase A f propagates to the structure 454 at the incident angle θ3, and the V ΊΓ θ3 is diffracted from the structure 454 and propagates through the detecting optical elements 460, 466 to the _ 器 · if the four ray 2 , each light _ _ way Track all the optical components passing through the optical measuring tool _ 昭 射 邻 测 测 测 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到 到The application and purpose of the thin section _, the number of lines side measurement, the first incident angle $ of the output illuminating light 436 on the fabric 454 of the working piece may be different for each ray. Then, , ^ ^ / / described ^ other factors will affect the propagation parameters of the arrival of the two] = like the incident angle, azimuth 'intensity of each optical component, and = = from the 5 Α疋 architecture diagram 'margin ray chase The vertical 500 utilizes a mirror-like reflection of the apricot wind _ cow's element having an optical axis 560 as shown. Mirror 550 has an entrance face 疋 15 201209371 This entrance face is the plane in which the input and output chief ray are interlaced. Using the three-ray mode, 'the input rays 51' and 512 having the broken section A1, on the incident surface 574, reflexive; become the output rays 514 and 516 having the differential section B1. In the other case, the input rays 520, 522 propagate to the reflective optical element 55G to form an output differential surface B2. However, the input rays 520, 522 are reflected outside of the entrance face 574 and are reflected as output rays 524, 526 having different differential sections B2. The differential section B2 can be smaller or larger than the differential section B1. In another case, the input light rays 530, 532 having the differential cross section A1 are reflected on the incident surface 574 into output light rays 534, 536 having a scoring cross section B1. Figure 5B is an architectural diagram showing ray tracing 600 with refractive optical element 68 (e.g., a focusing lens) and optical axis 616. The chief ray 6〇4 forms an incident surface with the optical axis 616. With the two-ray model, the point 656 of the input ray 612 in the incident surface enters the first major surface 660 and propagates through the refracting optical element 68 〇 and the second major surface 69 〇. The first angle of incidence 以往 in the direction of the previous focus 676 becomes Light 672 is output. The first main surface 660 of the other input ray 608 outside the entrance surface propagates from the focus 652 into the refractive optical element 680 through the refractive optical element 680, exiting from the second main surface 690, and the second angle of incidence 以往2 in the direction of the previous focus 652 Become the output light 668. The chief ray 604 enters the first major surface at point 650, propagates through the refractive optical element 680, and exits from the second major surface 69, becoming the output ray 664 at a third angle of incidence θ3. Figure 6 depicts an architectural diagram 500 showing ray tracing having a reflective optical element 570' and an optical axis 584' that involve a change in magnification or reduction of the output light over the beam section. The input ray 508', 512 having the differential section A1, reflected by the mirror 57, into the output ray 528', 532 having the differential section A2', converges at the focus 588. In another condition 'the input ray 520, 524 having the differential section B1, propagates onto the reflective optical element 570', and the reflected light rays 536', 540' having different differential sections B2 are converged, converge at the focus 588, . When perpendicular to the direction of travel of the chief ray (not shown), the differential sections A1, and ΒΓ are measured on the same plane, and the differential section A2, and B2 are measured on another plane perpendicular to the direction of travel of the chief ray (not shown). , at which time, the magnification m of the input rays 508 and 512', that is, % = , can be calculated, and for the input light

A

16 S 201209371 520', 524' is % = ¥. For geometric considerations, the magnification on the beam section is variably 'that' even if it is an input beam that equalizes the intensity distribution, the magnification produced by the optical element causes a change in the intensity of the beam. This is a redundant system artificial ί, ί produces measurement error. For example, suppose a beam of uniform intensity comes from a source that is focused on the structure after passing through a series of optical components. When the line with the differential section Α1 is tracked from the light source, the differential solid angle Ωι of the first ray can be calculated by the ray tracing algorithm when focused on the structure, so the angle intensity of the light focused on the structure is wj H line. The differential solid angle ω2 can be calculated by the ray tracing algorithm, ie. The angular intensity distribution is usually non-uniform, that is, because the geometric magnification is not uniform, that is, 'that' 2 or 岣 or Α is (15) where A is the differential stereoscopic opening of the self-focusing point on the wheeled beam target of the differential section A1. The angle, Ω2 is the differential solid angle 'Ai, A2 on the input beam target of the differential section A2 on the input beam target is the differential area of the two incident ray sections on the source aperture plane. Therefore, the angular magnification distribution is uneven, and the angular intensity distribution on the numerical aperture is uneven' and causes systematic errors in calculating the intensity and polarization of the diffracted signal. Fig. 6B is a plan view of Fig. 8 showing the ray tracing of the refractive optical element 804 (e.g., lens) and optical axis 820, which represents the enlargement and reduction of the cross section of the output light. Imagine a two-ray model containing two rays, two input rays 8〇8, 812 entering the refractive optical element 8〇4, traveling through the refractive optical element 8〇4 and the first main surface 844, and refracting the light rays 816, 822 to be directed toward the focus Point 840. In a different model, the two-ray model includes input rays 850, 854 'the two rays enter the refractive optical element 8〇4, travel through the refractive optical element 804 and the first major surface 844' to refract the light 830, 834 to exit toward the cluster Point 840. The second major face 824 is used for ray tracing and is used when different one or more rays & of the opposite direction pass through the refractive optical element. As mentioned in Fig. 6A, the light is on the beam section 17 201209371 There may be an enlargement or reduction of the output light. Light tracking can model the optical components, reflective or refractive components of an optical metrology tool, allowing light to change direction and properties as it enters, travels, and exits the optical component. Light 904 is refraction 998 998 _, / a * _ ^ a, & u a.

Ee疋 Input electromagnetic field with abnormally rounded light, its output Figure 7A is an architectural diagram 900 showing ray tracing in an optical component, indicating the strain birefringence of the output light. Birefringence occurs when a beam of light passes through a double-refraction object, the beam splits into a divergent beam, typically a non-abnormal (Ordjjjgjy) beam and an extraordinary beam. Strain birefringence results from external forces and/or deformations on materials such as stretched fibers, thin tantalum materials, or strains resulting from the adhesives used in the production of optical metrology tools. Considering a single input ray 9〇4, the electromagnetic field is Es and Ep, and the ray 904 enters the double refracting material 928 at an incident angle ρι with respect to the normal 912. The dual refractive material 928 has an optical axis 942 that is perpendicular to the plane of the paper. Input

What are the non-abnormal and abnormal rims?

(16) Available from: 201209371 Ί-Α ' is the imaginary unit 4 is the wavelength in vacuum, / is the material's j degree. Because of the phase delay, the polarization of light changes from the input state (slave + job) to =;, + exhaust. In general, for most optical components, it is previously intended to be used on the rotating surface, and in the parametric example of the dimming signal, the residual ray is Ε. Ee with different f rays. In this case, J number is η. With ne, therefore, after traveling a distance ^ within the material, experience a different phase delay, #-〇·/. As for the example of strain birefringence, birefringence is a function of the ship's volume and the pass f is not uniform across the wire. This uneven strain is the same polarization state after the birefringent material changes the polarization state of the light. This is called the generation of strain birefringence. It will be described from Chapters 9A and 9B that the polarization state of the output light can be changed by the input light of the sentence i. These changes are expressed as the phase, amplitude, and direction of the polarization state. This is more accurate for the travel in the system, and the ray of the present invention can be used to ignore the error caused by the source of error. 19 201209371 Figure 8纟t read 95G 'represents the ray of the refractive wire element in the same optical material as the output _ raw output' in jt Λ 968 J 968 ^ ^ ϊ ^ 954 rate ηι 'the refractive index of the second layer is & ^Specially different from the wavelength of light 958 in the first layer 954, ^^ kg 2 wavelength; ι=ϋ=7(£^1=1.; /〇(ω0/2τζ) η2} 0 C {ω0 /2π) (18) The rate of material U is the first speed of light: C is the speed of light in vacuum '" is the wavelength of the layer & When New Zealand's light ray traces the architecture diagram _ 'where the polarization state follows 9A1', the incident surface is the main light ray first axis 疋A out of the flat surface of the human face with a linear polarization material _8 is recorded Erzhi line · Two-way ^ is also linearly polarized. The incident surface has the same linear polarization outside the brain = tJ〇1G44f is broadcasted through the optical element 8 to become the ^10 line 1020 with the polarization state P3. The chief ray 1 〇 48 in the incident surface propagates into the output ray 1 〇 the output ray converges to be difficult to delete. Normally, the polarization (4) p3 is low because the light passes through the human face and the optical element surface isiii in the state of the non-uniform sentence. Polarization, biased by the artificial factors of "two 糸, · 克 钕 。. The electric field of the output light can be a linear polarization p2 such as 1028, or can be like a circular hopper of output light 1 〇 2 当 when the light wave electric field will rotate Participate in 9Α1, Ρι can be ^

S 20 201209371, often linear, p3 can be elliptical. Conversely, Ρι can be non-polarized, so that the light must be knives and become a positive polarized state, and each polarization state must be tracked separately. Orthogonally polarized light, the two methods - $ placed in the class n: according to the polarization state is off, partially coherent. Figure 9Α2 depicts the architecture without ray tracing. Figure 1〇〇〇, where the light is propagating through the refractive optics, the polarization state changes, and the refracted beam has no birefringence. Similar components have similar component symbols. The input light rays 1 〇 4 〇, 1 〇 44, and 1 〇 48 are polarized. The light transmitted through the surface of the optical element becomes the light ray 1012, 1028 having a polarization state π and the output light 1 〇 2 具 having a polarization state. Because of birefringence, the decay can be linearly polarized. Ρ2, Ρ3 can also be linearly polarized. Referring to FIG. 9A, the polarization state of the input ray 1564 is changed when reflected by the reflective element 1 into the output ray I536 in the incident surface, and the incident surface is defined by the chief ray (not shown as the optical axis 152G of the reflective element 1510. Similarly, # When the anti-Navigation component (10) is reflected out of the output light 1538, the polarization state of the input light 156() changes, such as from linear polarization 1568 to non-polarization 153. Figure 10A shows the light refraction and reflection in the film layer. The ray tracing frame 1300. The false wire is shown as a single-input light 13 〇 2 beam passing through the material M3 such as air at an oblique incident angle. The input ray 1302 is partially reflected at an angle % as the first reflected output ray 13G4, and at a refraction angle The portion 1 is refracted through the material M2 such as the wafer layer to become the first refracted ray 1314. The first refracted ray 1314 is further reflected in the material == lower boundary portion (four) to reflect the light 1316, and partially leaked over -

Ml becomes the refracted light 1326. The portion of the reflected light edge 1316 propagates with the second reflection ^ 1306 and is partially reflected at the upper boundary of the material M2 into the light 1318 to return to the material M2. The light 1318 repeats the partial reflection and refraction of the second process, and the light ray i3i8 is partially diffracted into the light 1328 through the simple M1, partially reflected into the light 132 and partially refracted through the material M1 of the next layer, and continues to be partially reflected. The light ray 1322 is to the material M2, part of the light ray 1330 enters the material M1, and part of the light ray becomes the third reflected output light 13G8. In the light reflection, refraction "repeating process towel" light electromagnetic energy can determine the reflection and propagation of the incoming light passing through the material layer. The following equation can be calculated: 201209371 Electromagnetic field E from the incident field, total reflection field, propagation field Composition, each field has two polarization states s and p, · (19) (20) (21)

Es.ou. =Γ5 Ep,out =fp-Ep . Total reflection intensity where ε is a constant number, rs and rp are complex reflection coefficients, and E is an electromagnetic field. FIG. 10B is a block diagram showing the ray tracing of multiple rays on a structure. The edge is shown in the first architectural diagram 1400 of the two-ray model. The input ray 1402, 1404, 1406 enters the structure L1 located above the first layer L2. The input rays 1402, 1404, and 1406 enter the structure at different incident angles θ^θγθ3, and are emitted as output rays 1412, 141, and 1408 at the angle of refraction. In the second architecture diagram 145〇 of the four-ray model, the input rays 1452, 1454, 1456, and 1458 enter the structure at different incident angles θι, %, %, and ,4, and are emitted at the refraction angles θ!, θ2, θ3, and Θ4. It becomes the output light 1466, 1464, 1462, and 1460. As shown in the figures and architecture diagrams, the number of lines in the ray tracing model can be one or more of the light required by the application. It should be noted that the computer resources required for simulation increase with the amount of light, and based on time constraints, computer resource requirements, and accuracy factors, an optimization program can be used to obtain the minimum amount of light. Figure 11A depicts a ray tracing architecture with refractive optical element 1504. Figure 158A depicts the scattering and stray light effects. In the two-ray model, the input ray 1528, 1532 enters the refracting optical element 15 〇 4 as a lens, and the light propagates through the optical element 1504 into refracted light 1550, 1552, which is when the refracting optical ploughs and bypasses the impurity 1544. The impurities 1544, the output rays 15〇8, 1512, such as the residue, which are located inside or on the surface of the refractive optical element 1504, have a smaller cross section due to convergence. In the case of the other two rays _, the input light 1582, 1540' surface defects 1548 (such as grooves or scratches or refractive optical elements 15 〇 4 surface non-reliance), the wealth of the refracting silk element 15G4 into ! 554, 2556 Input light 1582, 154 〇, produces an output light touch, 1524,

22 S 201209371 ^ Light $1584, 1524 when scattered outside gives a wider cross section. The output light 15 „, the diffused light outside the device 1516, may cause scattering shadowing of the measurement. = The light ray, such as the output light 1524, affects the measurement of the detector 1516. The input and the ray of the light are used to determine The direction of the light propagating to the detector 1516, and the interpretation of the structural profile intercepts the simulated model of the optical measuring device. Figure 11B shows an architectural diagram 1600 of the reflective optical element 1604 illustrating the scattering artifacts > Reflected by a defect 165〇 (such as a residue or contaminant on the inner surface of the reflective optical element 160^), a first output ray 1620 that is directed toward the optical metrology component 1630 is generated and reflected into diffuse light, (4) The 6 gamma second output ray 1628 is not reflected to the side device 1640, and the other input ray 1612 is reflected by the cavity 1654 (such as the irregular embossing on the surface of the concave or reflective optical element 16G4) to become the first -^ Output = 1616, and is reflected by component 1630 as diffuse ray 1636, reflected to detection = 40. Second output ray 1624 is reflected to the detection area outside detector 164 二 二 2 2 2 2 2 2 2 2 2 Light and The 1624, 162 出 affects your 164G measurement. The ray tracing of the input and output rays is used to illuminate the direction of the light of the two soft people and the detector. This data is used with other beam travel parameters. The model of the optical measurement tool is adapted to achieve the total accuracy requirement of the application. FIG. 12 depicts an exemplary flow chart 1700 showing the measurement output signal for sampling one or more contour parameters of the sample. The calculation method is to select the number of rays used to model the optical measuring tool in step 17 。 5. A light ray, such as a main line, may be selected. In other embodiments, two or more rays may be selected. In the case of _ demand. When selecting the quantity, the benchmark can be a secret structure or a similar historical modeled data and application or modeled data of the similar application. - 3 is the crystal structure of the ruthenium Shaped ^=resistance, in this example, used to model the number of ray initiations for similar applications in mG, for each system from the light-sample structure, the sample can be the pattern on the workpiece. Or unpatterned The structure may include a wafer structure such as a film, a grating or a repeating structure, a two-dimensional coil, a 23093; a 201209371 structure, or a three-dimensional structure. The wire travel parameters may include the following f-position angles, light intensity The orientation and cross-section intensity distribution are white. The polarization state, the uniformity of the polarization state change, the uniformity of each sentence, the pitting point, the transmittance and the reflectivity, the intensity propagation unevenness of the line, and the amplification of the light are not The hooks of the shadows === The change of the uneven polarization state of the first light can be determined by, for example, optical ^, "Α,, strain birefringence (Fig. 7Α, 7Β): due to 6Α'6Β)

(Fig. 9Α2, 10Α) The self-birefringence and strain birefringence can also be caused by the _transmittance and the reflective paste 1GA optical effect (Figs. 11A, 11B). And the first shot and the redirection. In step 1715, the selected beam propagation parameters are from each of the samples. In the _72 wipes, each will produce an age. There is a method of _ fixed silk strength 1 I textbook at this time Cai Lai people agency number "invention name" g gift ation toface f〇r ^ brother 7 =

System", the award date is June 20, 2006. In step + ° 〇 SY 溆 溆 low 柩 管 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , - or more ί·5: As described above, the contour capture system may use regression, database matching, or an illustrative block diagram of the 1800 system that uses the optical metrology tool brain and ray tracing methods to determine sample contour parameters. The optical system 1804 is calibrated using the specifications of the optical metrology manufacturer and the calibrator 1824 in the processor 1820 to generate calibration parameters. _ Processor lion and _ optical measurement (10) * and the optical measurement tools required for the application _ set the operation settings to produce the optical tool model body. The optical metrology tool model body includes the characterization of the illumination beam: the number of rays, the wire travel parameters, the fresh recording, and the like. The information 1806 from the optical measurement tool 18〇4 is transmitted from the optical measurement tool 18〇4 to the processor 182〇

24 S 201209371 adjuster lion. The signal adjuster deletes the optical measurement and expands the calibration parameters to rely on the _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ The round-robin (4) may be used by the regression module 1842, the 3 matching module 1844 and/or the machine learning system 1846 to determine the structural norm parameter 1832 for the processor 1820. The processor 182 uses the brain to change the optical measurement tool's brain's adjustable = such as Beixun's feedback beaker

The diagram illustrates an exemplary flow chart for the method of optimizing the number of rays and the beam travel parameters as well as the parameters of the hybrid configuration. In the number determination, one or more quasi-breakiness targets are used. Accuracy may include TMU, confidence zone _), standard uncertainty 2 _ set to (four) or lower, ί/ to indication' In another embodiment, the accuracy target indicates TMU disc C1n or lower. Set the CI to 90% or higher. In step ·, ^ is again 0.5 or lower, the number. As mentioned above, one or more rays can be used. Select the light to the light of the tool model f. Step 1915 ‘Based on the structure·the fixed number of alcohols, the parameters. The choice of beam travel parameters can be based on two 1 beam 仃 selectable beam travel parameters. If the setting is set to variable or fixed, the effect of the fixed value is negligible. The optical travel parameter setting is set to a fixed value. At the value of g, the parameter is based on the vendor's output measurement signal. In the step-by-step, the structure of the diffracted information, the adjustment of the rear wheel output measurement signal and the calibration of the optical measurement tool, the history of the poor material, the vending data, or the rim model Chemical. First, first learn the tool model and sample in step 1935, if the gentleman, the upper nine a, the total number of rays, the selected beam target, in step 1940, the number 'and repeat the adjustment after the measurement wheel Figure 2: Optimization of the 201209371 model and the contour model, comparison steps, and direct polarization. It is also assumed that - or multiple accuracy targets are slaves or more. After the fine and light, the amount of money will be taken to avoid the dung. ^ or · The reference work piece with the known number of rounds will be obtained. The blue and CI of the parameter are lower than the set 〇.50 2: the variability can be adjusted to include the polarization state, the scattering effect, and/or the diffuse light effect. Contour parameters, Ai sin is adjusted for variable or fixed contour parameters, 3 is changed to 〇5〇 or lower and 90% or higher (10). The kernel sequence is set up until an exemplary block diagram 1950 of Figure 15 Green Automatic Processing and Device Marriage is reached. In step 1955, the signal of the bi-parameter is used; the optical measuring tool can be used; the measuring device; the k-plane, or the degree of discrimination is set by JS i^o: ° t. , 〇冓 to ^ wheel corridor parameters modify at least - process parameters or equipment to determine, use the corridor parameters of the iiti^ii00" ^ 2000 should,, the vocal ίί, two | & in Figure 16 love is not connected - Manufacturing_, 2002, but before the first manufacturing ugly fiber, the k group can be used in the photoresist layer applied to the wafer _ light and / or development, #与旦>丨=2002 implementation. The optical measurement system toilet is similar to the light-drying system of Fig. 1, first 40. In an exemplary embodiment, the optical measurement system toilet includes

26 S 201209371 and processing 112G1G. The optical quantity station 2_ is used to measure the diffraction signal of ίϊ=ίϊ. The processor 诵 is used by the optical measuring tool to measure ' ί ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ , the reduction of the diffraction, the light that can be traced by light

In an exemplary embodiment, the optical measurement system 2_ may also include a plurality of values of - or multiple profile parameters. For example, when 2 is closed to the right of the simulated diffracted signal, the data is used by the database side to determine one or more contour parameters of the one or more contour parameters. - or multi-valued one or more - or multi-process parameters 2 to fabricate the cluster cluster fine 2 before or after processing the wafer. ^:^ group ^=6 = two system 2014. The wheeling parameter of the output 'training machine learning', although the exemplary embodiment is described above, but it is used for independent measuring and erecting 'self; pair; circle 27 3 201209371 form, therefore, the invention should not be limited to the above description DETAILED DESCRIPTION OF THE DRAWINGS [FIG. 1] FIG. 1 is an architectural diagram 'illustrated - an exemplary embodiment' in which light is used to determine the contour of a structure formed on a semiconductor wafer or substrate. The page 2 can be continued as an exemplary optical measurement system in accordance with an embodiment of the present invention. Figure 3 illustrates a prior art optical quantity, which is between the optical measurement tool and the processing module. Figure 5A shows an example of an optical measuring device using a ray tracing method. Figure 5A shows that the ray line using the ray of the reflecting optical element can enter the incident surface of the reflecting optical element, or at the incident surface. Into the face into the 1. Figure 5Β shows an architectural diagram using ray tracing with a refracting optical element that re-enters the entrance face of the refracting optic or enters beyond the entrance face. "Ding Xiaomi's diagram 6' shows the architecture diagram' indicates the use of ray tracing in a reflective optical element that involves the amplification of the output ray section. Figure 6Β depicts the architecture diagram' indicating the enlargement or reduction of the output ray section. The refracting optics use ray tracing. Figure 7Α shows the architectural diagram of ray tracing using optical components, showing the strain birefringence in the output ray. Figure 7Β shows the two-axis schematic architecture of the optical wavefront, including the electric field of non-abnormal light. Εο and the electric field Ee of the extraordinary light. Figure 8 is a block diagram showing the use of ray tracing in the refracting optical element, showing the change in the wavelength of the output ray. Figures 9A1, 9A2 show the structure of the polarization change as the light propagates through the refracting optical element. Figure 9B is a block diagram showing the change of polarization when the light is reflected by the reflective optical element. Figure 10A is a block diagram showing the ray refraction and reflection of the light in the film layer. Figure 10B shows the multiple light leaving structure. Schematic diagram of ray tracing. Figure 11A shows an architectural diagram using ray tracing with refractive optics to show scattering and diffuse light effects.

28 S 201209371 ′” scatter Figure 11B illustrates the architectural and non-emissive effects of the ray-tracing using reflective optical elements. .... ° or 争 示例 杂 杂 杂 杂 杂 , , , , , 杂 示例 示例 示例 示例 杂 杂 杂 杂 杂 杂 示例 杂 示例 杂 杂 杂 。 。 。 。 。郅之· , I3 draw slightly Wei Wei, table _ silk 6 method 1, determine the sample contour parameters of the system. /, Xingxian line tracking party ^14 draws an exemplary flow chart, showing the optimization of the beam travel parameters and the number of rays simultaneously with the optical number..., "charged" deconstructed contours Figure 15 depicts an exemplary flow Figure, shows the method used to insert and use the contour parameters. Columns of processing and equipment control Figure 16 shows an exemplary block diagram showing the method used to define and use the contour parameters. Judgment [Major component symbol description] 40 optical measurement system 41 measurement beam source 43 measurement illumination beam 45 incident angle 47 wafer / substrate 49 diffraction detection beam 51 measurement beam receiver 53 processor · 57 diffraction Signal 59 Sample Structure 60 Simulator 100 Optical Measurement System 101 Wafer 102 Platform 103 Platform Base 104 Wafer Alignment Sensor 29 201209371 105 Light System 106 Optical Output 110 Illuminator Sub System 111 Optical Output 115 Selector Sub System 116 signal 120 beam generator subsystem 121 output 125 reference subsystem 126 reference output 130 first optional reflection subsystem 131 first path 132 second path 135 low angle focus subsystem 136 output 140 A reflection subsystem 141 output 145 high angle focus subsystem 146 output 150 second reflection subsystem 151 output 152 output 155 high angle collection subsystem 156 output 160 second optional reflection subsystem 161 output 162 output 165 low angle collection subsystem 166 output 170 analyzer subsystem

30 S 201209371 171 Output 175 Measurement System 180 Camera Sub System 182 Development Sub System 184 Illumination Sub System 186 Output 190 Auto Focus Sub System 195 System 304 Optical Measurement Tool 308 Universal Interface 312 Processing Module 400 Optical Measurement Tool 404 light source 408 first illumination ray 416 second illumination ray 418 center light, line 420 slice 422 light 424 output light 428 light 426 illuminate optical element 430 focus element 432 output light 436 output light 440 output light 444 motion control system 452 work piece 454 Structure 456 Detecting Light 458 Detecting Light 201209371 460 Collecting Optical Element 462 Detecting Light 464 Detecting Light 466 Collecting Optical Element 468 Light 470 Output Light 472 Cut Surface 476 Detecting Light 478 Output Light 480 Detecting Light 484 Detector 490 Illumination portion 494 detection portion 500 ray tracing 500' architecture diagram 508', 512' input ray 510, 512 input ray 514, 516 output ray 520 > 522 input ray 520', 524 'input ray 524, 526 output ray 528' 532' output light 530, 532 input light 534, 536 output light 536' 540 '550 output light reflective optical element 560 axis 570' of the reflective optical element / focus mirror 574 incident surface 580

S 32 201209371 584' optical axis 588' focus 600 ray tracing 604 main ray 608 input ray 612 input ray 616 optical axis 650 point 652 focus 656 point 660 first main surface 664 output light 668 output light 672 emit light 676 focus 680 refractive optical Element 690 Second Main Face 800 Architecture 804 Refracting Optical Element 808, 812 Input Light 816, 822 Refracted Light 820 Optical Axis 824 Second Main Face 830, 834 Refracted Light 840 Focus Point 844 First Main Face 850, 854 Input Light 900 Architecture Figure 904 Input Ray 912 Normal 201209371 924 Refracted Light 928 Double Refractive Material 934 Non-Exceptional Output Light 938 Abnormal Output Light 942 Optical Axis 950 Architecture Figure 954 First Layer 958 Monochromatic Beam 962 Second Layer 964 Light 968 Object 1000 Architecture 1000' architecture diagram 1008 optical component 1012 output light 1020 output light 1024 focus 1028 output light 1040 input light 1044 input light 1048 main light 1050 incident surface 1200 architecture diagram 1300 architecture diagram 1302 input ray 1304 output ray 1306 second reflection output ray 1308 Three-reflected output light 1314 first refracted light 1316 reflects light 201209371 1318 ray 1320 ray 1322 ray · 1326 refracted light 1328 ray 1330 ray 1400 architecture diagram 1402, 1404, 1406 input ray 1408, 1410, 1412 output ray 1450 architecture diagram 1452, 1454, 1456, 1458 input ray 1460, 1462, 1464, 1466 Output Light 1500 Architecture Figure 1504 Refracting Optics 1508 Output Light 1510 Reflecting Element 1512 Output Light 1516 Detector 1520 Optical Axis 1524 Output Light 1528 Input Light 1530 Unpolarized 1532 Input Light 1536 Output Light 1538 Output Light 1540 Input Light 1542 Elliptical Polarization 1544 impurity 1548 surface defect 1550 refracted light 35 201209371 1552 refracted light 1554 refracted light. 1556 refracted light 1560 input light 1564 input light 1568 linear polarization 1580 architecture Figure 1582 input light 1584 output light 1600 architecture Figure 1604 reflective optics 1608 input ray 1612 input Light 1616 first output light 1620 first output light 1624 second output light 1628 second output light 1630 element 1632 diffused light 1636 diffused light 1640 detector 1650 defect 1654 cavity 1700 flow chart 1705, 1710, 1715, 1720 1725,1730 step 1800 the optical metrology tool system 1804 1806 1808 reserved data information processor 1820

S 36 201209371 1822 optical measurement tool model 1824 calibrator 1826 signal adjuster 1830 adjusted measurement output signal 1832 contour parameter 1840 contour extractor 1842 regression module 1844 database matching module 1846 machine learning system 1900 flow chart 1905, 1910, 1915, 1920, 1925, 1930, 1935, 1940 Step 1950 Block Diagram 1955, 1960, 1965, 1970 Step 2000 System 2100 Block Diagram 2002 First Manufacturing Cluster 2004 Optical Measurement System 2006 Second Manufacturing Cluster 2008 Optical Volume Testing Tools 2010 Processor 2012 Database 2014 Machine Learning System 2016 Measuring Processor 37

Claims (1)

201209371 VII. Patent application scope: ^ ί 作 作 上 上 上 上 上 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括 包括The optical measuring tool and the second iim, the contour model has a sample contour parameter; the light has a beam first beam, an optical element, a beam inspection, a detection method, the method quantity, and a volume model The number of rays of the illumination beam (b) selects the number of beam travel tables of the wire measurement aid model; (C) uses the processor: 'The selected number of light-structures of the red-haired woman-shaped structure Wei Jia 11 (4) calculating the intensity of each of the diffracted beams and the total intensity of the diffracted beam on the side device, and calculating the cross-section of the diffracted beam, and calculating the (c5) of the diffracted beam from the total intensity and the polarization calculation output. The utility uses the measured wheeling signal, the calibration data of the optical quantity, and the rounding algorithm to retrieve the one or more rim parameters. The capital of 11 ^ system bullying. On the sample model == number database matching or machine learning system, perform simulation calculations to produce the original. If you apply for the model (4), use the optical quantity system to determine the workpiece loading S 38 201209371 A method for contouring a parameter, the number of rays of the optical measuring fixture model, the one or more rays, the illumination beam comprising one or more illumination beams and/or wherein the detector comprises two or more a detector, and the workpiece is a wafer or a substrate. 4. The method of applying the second aspect of the patent scope to the contour parameter of the structure, the wire travel parameters including one or more of the following corners, the human face, and the orientation of the light. , beam intensity, strong cross-section and sentence uniformity of the knife cloth, polarization state, uniformity of polarization state change, sex and reflection effects, light scattering effects and/or diffused light effects. 5. If the wire measurement system of the fourth application of the patent scope determines the profile parameters of the sample structure on the workpiece, the polarization state of each light of the towel may vary depending on the plane into which the light is entered, and The difference in the change in polarization state is included in the calculation of the diffracted signal modulus. The beam intensity of the sample in which the sample is taken is different when passing through the optical elements, and the difference in the intensity of the eight light numbers is included in the simulation calculation of the diffraction signal; and/or ^ 'Each light_beam intensity is distributed in the cross-section mode 1 of the light passing through each optical element. The difference in beam intensity distribution is included in the diffraction signal, and the sample is made on the wire 4 of the diffraction signal. The method of the number 'the towel is unevenly defined because of the following one or (8) uneven transmission of light from the source; (b) uneven propagation of light passing through the optical element; (C) Irradiation beam and the diffracted beam are not uniformly magnified; 39 201209371 8 · A method for determining a profile parameter of a sample structure on a workpiece using an optical measurement system according to claim 5, wherein the polarization state of the light changes over the ray section The difference can be different, and the difference in the polarization state change is included in the diffraction signal simulation calculation. 9. A method for determining the profile parameters of a sample structure on a workpiece using an optical metrology system according to item 5 of the patent application, wherein the polarization state change of the light can be caused by: (a) strain birefringence due to inclusion of the optical element And/or flaws in the ridge and the uneven polarization state change across the ray section; (b) the unpolarized state change of the cross-beam section due to the incident surface changed by the optical element; and/or one (c) Light scattering due to stagnation or defects of optical components. For example, the method for determining the contour parameters of the structure on the workpiece using the optical measurement system in the second paragraph of the patent scope: 'wavelength 3 quality calculation to generate the age-of-the-money silk attribute is not specified for The optical setting of the silk 4 measuring device and the geometrical parameters of the geometrical optical juice including the reflection angle, the diffraction angle, and the diffraction angle; the polarization parameter of the 1 line is in the _(丨嶋) matrix or the array is not, It is stored for each light and wavelength of the selected light. ====Items _Miscellaneous _ Workpiece loading does not = Shooting ===== Fine two or more 弁 / ί^ί1 Transfer from the optical setting and geometric parameters of the optical measuring tool and the first In the consideration of the behavior of Lai and physical optics, calculate the polarization of each light. 3 Pairs of 201209371 Shaped mechanical parts are loaded with the results of the filaments. The parameters include the position of the optical components, optics. The geometry of the component first dissipates the inhomogeneities or defects present in the component; and/or wherein the calibration parameters include strain birefringence, scattering, and/or diffuse light. Using the optical measurement system to determine the sample of the workpiece on the workpiece, the optical measurement system is recorded to achieve I =: number = the sample structure has a contour parameter, and the definition of the inclusion is determined by the number " = 萼 萼: = the sample structure - or a plurality of 誃 誃 誃 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ a selected beam travel parameter, the = amount; =, a signal adjuster for utilizing the optical metrology tool model and the optical quantity structure: the sample or the plurality of rim parameters; the operation module 'determines the The sample structure is optimized, and the contour model simultaneously performs the best simulation g. 'One or more of the pulleys of the wheel picker and includes the diffracted signal 201209371 Apply for the special 13th item · Optical development work piece The system of contour parameters of the structure: ”The heart is loaded, wherein the selected number of rays is one or more rays; the instrumental iLSS11 includes a regression module, a database matching module and/or a mechanism, wherein the sample structure is a semiconductor Grating or crystal Or a repeating knot on the substrate. 15. For example, the contour parameter of the optical system is the same as the contour parameter of the optical system, wherein the beam travel parameters include the following: 2, 2, azimuth, and human shooting. Orientation, intensity, section strength and knife uniformity, polarization state, uniformity of polarization state change, sex and reflection effects, light scattering effect and/or diffused light effect. Item 13 of the system for utilizing optical quantity to determine the profile parameter of the structure of the workpiece, the towel adjuster is further produced by using different optical lines of the optical tool to enter different planes of the optical tool. The difference in the polarization state of each light in the selected number of rays. 17. For the system of applying the sub-paragraph 13 of the optical measurement secret profile of the structure: 7 where the signal regulator is used to consider a difference in beam intensity distribution across a ray section of each of the plurality of optical elements, and wherein a difference in distribution is included in the simulation calculation of the diffracted signal; wherein the signal adjuster is further used The amount of light in each of the selected rays is different from the light's deviation. ^ Where the depolarization difference is included in the diffraction signal, the difference is 1f, and/or the age adjuster is used to consider the selected light. The depolarization of each light in the quantity due to: (a) the variation of the strain birefringence and/or fatigue of the optical element and the variation of the polarization state of the cross-ray section caused by the sputum; 42 S 201209371 The change of state; and the unevenness of the trans-lighted surface caused by the surface of the person who changed the & (C) is caused by the shouting or observation of the optical component. The optical measurement is loaded by the two-dimensional detector. Further, it is used to consider the wavelength variation when the light of each selected light propagates through a different refractive index material layer, and wherein Ϊ wavelength Aihua is included in the simulation calculation of the diffraction signal; - ΐη the signal adjuster is used more Considering the angle of incidence, the azimuth angle, the slitting and the number of light, the number of the light, the number of light, the number of light, the number of light, the number of light, the number of light, the number of light Qian Zhi 11 is more ribbed to take the proceeds of the optical measurement tool V, number and the secret design parameters (4) determine the #geometry parameters and polarization parameters. Wire-shaped button-shaped work piece refraction =====Qing withdrawal test 嶋 strain double 20 kinds of work pieces on the sample structure of the wheel corridor parameters of the system includes: System ^ Ϊ ί ^ measurement system is optimized Achieved - or more accurate pace of the optical workmanship reduction, the number of selected samples and the selected light sample formation model, 5 Hai learn! The measuring tool is used to measure one or more accuracy targets of the number of the measuring instrument leaving the number, the IS measuring tool and the measured diffraction signal. 21. If the application of the optical measurement system determines the top of the work piece, the sample structure is the second part of the patent range. A system for determining a profile parameter of a workpiece by using an optical measurement system, wherein the beam travel parameters include one or more of the following, azimuth, human face, direction of light, intensity, and section strength. The optical quantity of the 2G term (4) The system of loading the i:: parameter, where the – or more quasi-Wei goals include the ί confidence interval, the standard uncertainty, the combined standard uncertainty, and / == = Item of the general 幽 幽 作 作 作 上 ί = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = , with optical measurement tools 忿 该 该 该 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ... a + component, a diffracted beam, a measuring Si measuring tool to measure the burning signal leaving the sample structure = the measured output money and calibration data, determine the adjusted measurement and use the adjustment Post-measurement output signal and the parameter drowning algorithm Optimizing the contour model; first, if the one or more accuracy targets are achieved, adjusting the selected beam travel parameters and/or the contour parameters, and the parameter and the algorithm, the money determines the 5S model and The contour model is optimized until it is reached; more 1s are closed (four) and the step is 2 = 2 (the _ woven (four) position signal step 2 = 2; with the light source to the sample smooth line ϊ ί ί The light of each of the rays of the hybrid structure to the selected light of the Gu 11 is the Wei polarization of the light of the diffracted beam on the detective; the vibration; and especially the surface of the lens produces the total intensity and deviation of the diffracted beam The signal is measured from the total intensity and polarization calculation. 2 The slave's fine-grained green-sensitive readings are loaded with angles: man's angle of shooting, azimuth difficulty, uniformity of cross-sectional intensity distribution, polarization 45 201209371, cancerous, polarization-stricken changes, and/or (four) Should be; j is more 撷 = method to take the one or the system, perform simulation calculation = = library matching and / or machine engineering use of the system to determine the workpiece after the sample is off the measurement round signal The steps consider the polarization of each light == according to the first line through the optical measurement tool, each optical compensation, each signal considers the beam intensity distribution of each light in the step of wearing the light: the depolarization number of the light In the calculation; the difference between the depolarization and the depolarization difference is included in the diffraction signal. The depolarization of the light is considered to be due to the step of turning out the signal and its ίί; ίί^2 ίί^ JL·以骤::, specifies the reflection angle, the green f forest light material calculation includes, and the step of determining the adjusted measurement wheel output signal is determined by using the optical quantity S 46 201209371 ί 11 The resulting records, what parameters and polarization are determined Shoot, heart, 纟 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ , the confidence interval, the standard is not a monument ς ^ including total extension uncertainty and / or total measurement uncertainty. No, 'uncertainty, eight, schema: 47