CN104317168B - Mirror finish error analysis method relevant with parasitic light in extreme ultraviolet etching system - Google Patents

Abstract

Mirror finish error analysis method relevant with parasitic light in extreme ultraviolet etching system, belong to optical design simulation analysis field, solve the expansion intermediate frequency roughness be closely related with parasitic light in extreme ultraviolet etching system and absorb the problem causing mirror finish error analysis lower without unified measurement standard or method for solving.The method is: the minimum value of initial dark island size and maximal value, gets its mean value and enters lithography simulation for initial dark island size, judges that the relation of stray light level and emulation parasitic light threshold value is until both are equal, obtains the best secretly island size; The space periodic of expansion intermediate frequency roughness is Λ, corresponding spatial frequency , grating equation Λ sin θ=λ, λ is system exposure wavelength, the spatial frequency that the intermediate frequency roughness that is expanded is corresponding: , θ and θ ' is for incident light is expanding angle of diffraction ultimate value when intermediate frequency roughness occurs diffraction, and θ < θ '.The present invention can exactly for the work such as lithography performance analysis, the refine of subsequent optical element provide important evidence.

Description

Mirror finish error analysis method relevant with parasitic light in extreme ultraviolet etching system

Technical field

The invention belongs to optical design simulation analysis technical field, be specifically related to mirror finish error analysis method relevant with parasitic light in a kind of extreme ultraviolet etching system.

Background technology

Extreme ultraviolet lithography (EUVL) adopts the extreme ultraviolet (EUV) of 13.5nm as one of Next Generation Lithography (NGL) of operation wavelength, is the gordian technique realizing 22nm and following technology node thereof.Due to the exposure wavelength that it is ultrashort, system imaging, except will meeting minimum wave aberration (~ 1nmRMS), also needs to take into full account the luminous energy scattering caused by optical element processing roughness.According to scattering theory, the luminous energy scattering caused by optical element mismachining tolerance mainly can be divided three classes, low frequency roughness (LowSpatialFrequencyRoughness respectively, LSFR) the very small angles scattering caused, basic aberration is introduced, so-called optical element profile (Figure) in system; By intermediate frequency roughness (MiddleSpatialFrequencyRoughness, MSFR) small angle scattering caused, make luminous energy scattering in whole visual field, image planes are caused to form a uniform background illumination, thus reduce aerial image contrast, by being called parasitic light (Flare); The large angle scattering caused by high frequency roughness (HighSpatialFrequencyRoughness, HSFR) will make luminous energy overflow image planes, cause energy to decline, and reduce the output of system.

Parasitic light in extreme ultraviolet etching system is different from the parasitic light in deep UV (ultraviolet light) etching system, it can be considered to only to be caused by optical element mismachining tolerance, and the impact of the mismachining tolerance of different space frequency on system differs, the luminous energy scattering only having specific frequency domain mismachining tolerance to cause just can introduce a uniform background illumination to image planes, form parasitic light, reduce aerial image contrast, reduce photoetching process window ara, thus increase technology difficulty or cause photoetching process to realize, and system assignment is to the tolerance of parasitic light, mismachining tolerance frequency domain relevant with parasitic light in optical element mismachining tolerance and the parasitic light quantity of introducing all need by accurate analysis and measurements in emulation in early stage and later stage measure, for optical element processing provides guidance.Can the parasitic light wherein introduced by intermediate frequency roughness (MSFR) directly will affect photoetching and realize, and therefore must carry out Accurate Analysis assessment.

At present, the expansion intermediate frequency roughness be closely related with parasitic light is absorbed there is no and is sought unity of standard or method, homologous series extreme ultraviolet exposure experiments device Analysis for Stray Light is caused not possess theoretic continuity and consistance, as the micro-visual field extreme ultraviolet photolithographic exposure system (MicroExposureTools of three covers in U.S. LLNL laboratory, MET), although three cover systems structurally all have employed Swarzschild two mirror structure, functionally also there is consistance and hand down, but it is different to the definition of system parasitic light, and at home, the National University of Defense technology utilizes wavelet transformation optical element mismachining tolerance constantly to be decomposed in frequency field, and according to different frequency territory error, system Analysis for Stray Light is carried out to the different-effect of aerial image quality, but should often carry out a wavelet transformation based on the error frequency range analytical approach of wavelet transformation, its data point will reduce half, this just causes subsequent analysis data obviously not enough, lose confidence level.

At present, extreme ultraviolet etching system stray light level many employings Kirk method is measured in industry, namely the dark island (on high reverse--bias formula mask plate an isolated absorption region) having certain size to exposes, after the photoetching processes such as exposure imaging, the ratio washing clear zone and exposure dose that dark space uses completely is system stray light level, but in the selection of this dark island size, unified standard or method for solving is not formed in current industry, Stray Light Test result height is caused to depend on its size Selection, and at present this size Selection and the process node size that will realize without obvious relation between persistence.

Summary of the invention

Absorbing to solve the expansion intermediate frequency roughness be closely related with parasitic light in extreme ultraviolet etching system the problem causing mirror finish error analysis lower without unified measurement standard or method for solving, the invention provides mirror finish error analysis method relevant with parasitic light in a kind of extreme ultraviolet etching system.

The technical scheme that the present invention adopts for technical solution problem is as follows:

Mirror finish error analysis method relevant with parasitic light in extreme ultraviolet etching system, condition and the step of the method are as follows:

Solving of step one, best dark island size

(1) emulation parasitic light threshold value is determined;

(2) minimum value of dark island size is determined according to photoetching principle λ is exposure wavelength, and NA is the image-side numerical aperture of projection exposure system therefor, and the minimum value of this dark island size is the limiting resolution of projection exposure system therefor, namely all can not successful Imagewise exposure under any condition;

(3) the maximal value d of initial dark island size _max=D, the maximal value of this dark island size can the aerial image on Shi An island be by force 0 in any condition Xia Qi center area light;

(4) minimum value of Qu An island size and the mean value of maximal value first time lithography simulation is entered as initial dark island size;

(5) stray light level is obtained according in Kirk method about the definition of system parasitic light and lithography simulation aerial image, judge stray light level and the relation emulating parasitic light threshold value, if stray light level is greater than emulation parasitic light threshold value, illustrates that initially secretly island size d is too small, then by d assignment in d _minand enter simulation calculation next time, if stray light level is less than emulation parasitic light threshold value, illustrates that initially secretly island size d is excessive, then by d assignment in d _maxand enter simulation calculation next time;

(6) repeat step (5), final d will converge on a certain value d _island, until stray light level is equal with emulation parasitic light threshold value, i.e. d=d _island, then d _islandfor the best dark island size;

Step 2, utilization best dark island size solve spatial frequency corresponding to the expansion intermediate frequency roughness relevant with parasitic light

(1) the dark island of optimum dimension forms dark island aerial image after projection exposure system therefor, and clear zone forms clear zone aerial image after projection exposure system therefor, because there is the expansion intermediate frequency roughness relevant with parasitic light in projection exposure system therefor, incident light generation diffraction is caused to produce 0 grade of light and two bundle secondary light, ignore the low high diffracting grade of energy level, wherein a branch of secondary light enters into dark island aerial image, and superposition one bright background, forms parasitic light;

(2) dark island aerial image place image planes and aerial image place, clear zone image planes are at grade, if the distance between this plane and projection exposure system therefor bore is l ', optimum dimension dark island place object plane and place, clear zone object plane are at grade, if the distance between this plane and projection exposure system therefor bore is l, the space periodic of expansion intermediate frequency roughness is Λ, corresponding spatial frequency best dark island is of a size of d _island, system magnification is α, then in conjunction with grating equation Λ sin θ=λ, wherein, λ is system exposure wavelength, solves spatial frequency corresponding to the expansion intermediate frequency roughness relevant with parasitic light from two aspects:

1. from object space picked-up expansion intermediate frequency roughness:

$\frac{d_{\mathrm{island}}/2}{l}=\tan \mathrm{\&theta;}$

$\frac{\mathrm{FOV}/2}{l}=\tan {\mathrm{\&theta;}}^{\&prime;}$

2. from image space picked-up expansion intermediate frequency roughness:

$\frac{\mathrm{\&alpha;}{d}_{\mathrm{island}}/2}{l^{\&prime;}}=\frac{d_{\mathrm{island}/2}}{l}=\tan \mathrm{\&theta;}$

$\frac{\mathrm{\&alpha;}\&CenterDot;{\mathrm{FOV}}_{\mathrm{image}}/2}{l^{\&prime;}}=\frac{\mathrm{FOV}/2}{l}=\tan {\mathrm{\&theta;}}^{\&prime;}$

Wherein, θ and θ ' is for incident light is expanding angle of diffraction ultimate value when intermediate frequency roughness occurs diffraction, and θ < θ ', owing to expansion intermediate frequency roughness having the mismachining tolerance of different frequency, therefore actual angle of diffraction is the arbitrary value between θ and θ ', then the spatial frequency distribution scope that relevant with parasitic light expansion intermediate frequency roughness is corresponding is:

$\frac{\sin \mathrm{\&theta;}}{\mathrm{\&lambda;}}<f<\frac{{\sin \mathrm{\&theta;}}^{\&prime;}}{\mathrm{\&lambda;}}$

No matter be solve or solve from image space conclusion from object space unanimously, therefore can solve expansion intermediate frequency roughness respectively from object space and image space.

Further comprising the steps of before carry out step one: the theoretical analysis that best dark island size proposes

When a dark island of small size forms the first aerial image after preferred view exposure system, diffraction is there is due to the restriction of system value aperture, thus make the central area of first aerial image on the dark island of small size have light intensity, even use this small size dark island measuring system parasitic light, for preferred view exposure system, its stray light level recorded is not 0 yet, and this contradicts with the preferred view exposure system of hypothesis, illustrates that this small size dark island size Selection is improper;

When a dark island of large scale forms second space picture after more than half real projection exposure system, half real projection exposure system is made up of the expansion intermediate frequency roughness of preferred view exposure system, basic aberration and introducing parasitic light, the central area light intensity of the second space picture on the dark island of large scale is made to be 0, namely according to this large scale dark island measuring system parasitic light, for half real projection exposure system, its stray light level recorded is 0, this contradicts with half real projection exposure system of hypothesis, illustrates that this large scale dark island size Selection is improper;

Based on the consideration of above two kinds of situations, a dark island of optimum dimension is certainly existed between the dark island of small size and the dark island of large scale, when a dark island of optimum dimension forms the 3rd aerial image after projection exposure system therefor, projection exposure system therefor is made up of preferred view exposure system and basic aberration, the central area light intensity of the 3rd aerial image on the dark island of optimum dimension is made to be just 0, this just matches with the projection exposure system therefor of hypothesis, and emulates with utilization the systems compliant carrying out using in the process of parasitic light Tolerance assignment;

Two aspects are measured theoretical in conjunction with etching system emulation and system parasitic light, obtain the conclusion certainly existing the optimum dark island size that measures for parasitic light, instead of choose arbitrarily, make parasitic light become a kind of intrinsic property of system, relevant with lighting condition, optical element level of processing.

The invention has the beneficial effects as follows:

The present invention is intended in conjunction with extreme ultraviolet etching system parasitic light fundamental measurement principle and lithography simulation analysis principle, a kind of unified method is proposed theoretically in order to absorb the expansion intermediate frequency roughness relevant with system parasitic light, from the process data prognoses system parasitic light of optical element, have more directive significance to make emulation.

The present invention is based on Kirk method fundamental measurement principle and lithography simulation, solve and obtain utilizing Kirk method to measure the best dark island size of extreme ultraviolet etching system parasitic light, utilize and solve the best obtained dark island size, in conjunction with etching system topology layout, there is the mismachining tolerance frequency domain of light in each optical element with system parasitic light in quick capturing system, can exactly for the work such as lithography performance analysis, the refine of subsequent optical element provide foundation.The measurement that the method can be system parasitic light in industry provides a unified normative reference, is convenient to communication in industry.

Accompanying drawing explanation

Fig. 1 is the principle schematic that the dark island of small-medium size of the present invention forms the first aerial image after desirable projection exposure system therefor.

Fig. 2 is the principle schematic that in the present invention, the dark island of large scale forms second space picture after half real projection exposure system.

Fig. 3 is the best dark island size resolution principle schematic diagram for measuring parasitic light in the present invention.

Fig. 4 solves schematic flow sheet for the best dark island size measuring parasitic light in the present invention.

Fig. 5 is the principle schematic utilizing best dark island size to solve spatial frequency corresponding to the expansion intermediate frequency roughness relevant with parasitic light.

In figure: 1, the dark island of small size, 2, the dark island of large scale, 3, the dark island of optimum dimension, 4, preferred view exposure system, 5, half real projection exposure system, 6, projection exposure system therefor, 7, the first aerial image, 8, second space picture, the 9, the 3rd aerial image, 10, basic aberration, 11, expansion intermediate frequency roughness, 12, single mirror, 13, dark island aerial image, 14, clear zone, 15, clear zone aerial image, 16, incident light, 17, secondary light, 18,0 grades of light.

Embodiment

Below in conjunction with accompanying drawing, the present invention is described in further detail.

Mirror finish error analysis method relevant with parasitic light in extreme ultraviolet etching system of the present invention, condition and the step of the method are as follows:

The theoretical analysis that step one, best dark island size propose

In extreme ultraviolet etching system, parasitic light is considered to a uniform background illumination in Image space imaging, Image space image contrast will be caused to decline, thus reduce photoetching process window, in a photolithographic process, how much such system can tolerate background illumination, is that system assignment is to the tolerance of parasitic light.

In extreme ultraviolet etching system, the exposure of successive doses is carried out to the dark space of a specific dimensions on high reverse--bias formula mask, obtaining stray light level by analyzing its photoresist picture formed, the dark space of this specific dimensions being called island, being translated into dark island at this.

As shown in Figure 1, when a dark island 1 of small size forms the first aerial image 7 after preferred view exposure system 4, diffraction is there is due to the restriction of system value aperture, thus make the central area of first aerial image 7 on the dark island 1 of small size have light intensity, even use this small size dark island 1 measuring system parasitic light, for preferred view exposure system 4, its stray light level recorded is not 0 yet, this contradicts with the preferred view exposure system 4 of hypothesis, illustrates that this small size dark island 1 size Selection is improper, as shown in Figure 2, when a dark island of large scale 2 forms second space as 8 after more than half real projection exposure system 5, half real projection exposure system 5 is by preferred view exposure system 4, the expansion intermediate frequency roughness 11 of basis aberration 10 and introducing parasitic light forms, make the second space on the dark island 2 of large scale as 8 central area light intensity be 0, namely according to this large scale dark island 2 measuring system parasitic light, for half real projection exposure system 5, its stray light level recorded is 0, this contradicts with half real projection exposure system 5 of hypothesis, illustrate that this large scale dark island 2 size Selection is improper.

Based on the consideration of above two kinds of situations, a dark island 3 of optimum dimension is certainly existed between the dark island of small size 1 and the dark island of large scale 2, as shown in Figure 3, when a dark island of optimum dimension 3 is under specific lighting condition, the 3rd aerial image 9 is formed after projection exposure system therefor 6, projection exposure system therefor 6 is made up of preferred view exposure system 4 and basic aberration 10, the central area light intensity of the 3rd aerial image 9 on the dark island 3 of optimum dimension is made to be just 0, this just matches with the projection exposure system therefor 6 of hypothesis, and emulate with utilization the systems compliant carrying out using in the process of parasitic light Tolerance assignment.Only have utilize the stray light level of the dark island size of this best and the actual measurement of Kirk method could with early stage parasitic light distribution tolerance match.

Two aspects are measured theoretical in conjunction with etching system emulation and system parasitic light, obtain the conclusion certainly existing the optimum dark island size that measures for parasitic light, instead of choose arbitrarily, parasitic light is made to become a kind of intrinsic property of system, relevant with lighting condition, optical element process water equality factor.

Solving of step 2, best dark island size

Solving of best dark island size completes mainly through lithography simulation, in lithography simulation process, numerical aperture and basic aberration 10 is added in system, make diffraction that in the tolerance of parasitic light, complete removal system numerical aperture causes and the deflection of light effect that basic aberration 10 causes, and then recycling is based on the parasitic light Tolerance assignment result that accurately must could utilize emulation gained in the parasitic light experiments of measuring of photoresist using emulation as standard, namely in the experiments of measuring of parasitic light, the experiment utilizing Kirk method to measure parasitic light is refered in particular at this, must effectively the restriction of removal system numerical aperture and the luminous energy introduced of basic aberration 10 redistribute, guarantee in measuring process, the exposure of dark space, center photoresist must and only introduced by parasitic light.

According to lighting condition in extreme ultraviolet etching system, the established conditions such as actual measurement aberration, with Airy disk, (pointolite is by perfect lens imaging, central authorities are bright former spots, there is the light and dark concentric annular striped that a group more weak around, the central bright spot with first dark ring being wherein boundary is called Airy disk, its size determines telescopical resolution characteristic) be of a size of dark island size initial value, with the minimax strength ratio of aerial image in emulation for criterion, iterative algorithm is utilized to solve best dark island size, make the minimax strength ratio of aerial image in emulating far below a certain parasitic light measuring accuracy, namely making the minimum intensity of light of dark island aerial image 13 be just 0 (should be just 0 in theory, but due to simulation calculation error in actual solution procedure, the restrictions such as system parasitic light measuring accuracy, as long as reach a threshold value).

As shown in Figure 4, when calculating beginning, according to parasitic light measuring accuracy, simulation calculation error etc. determines emulation parasitic light threshold value, emulation parasitic light threshold value is determined by practical application request and software and hardware condition, as parasitic light measuring accuracy, current bibliographical information is all as 5%, 10%, 16% such numerical value, therefore 0.1% need not be set to, the round-off error that and for example in emulation, software calculates, calculating cellular size in Electromagnetic Calculation, if electromagnetic-field simulation is 2.5nm computing unit, so just do not need emulation to be accurate to below 2.5nm precision, while waste computational resource, excessive demand be it is also proposed to processing, emulation parasitic light threshold value in simulations theory is 0, not talkative emulation parasitic light threshold value is 0, because the final purpose of emulation measures in order to the parasitic light realized in real system, natively need to consider real demand, therefore can not replace by desirable threshold value 0.The minimum value of dark island size is determined according to photoetching principle wherein λ is exposure wavelength, and NA is the image-side numerical aperture of projection exposure system therefor 6, and the minimum value of this dark island size is the limiting resolution of projection exposure system therefor 6, namely all can not successful Imagewise exposure under any condition; The maximal value d of initial dark island size _min=D, the maximal value of this dark island size can the aerial image on Shi An island be by force 0 in any condition Xia Qi center area light; The minimum value of Qu An island size and the mean value of maximal value enter first time lithography simulation as initial dark island size, this emulation initial conditions comprise lighting condition, lithography projection exposure objective lens aberration etc. a series of measure determine, can all factors (but not comprising parasitic light) of the Realization of Simulation; Stray light level is obtained about the definition of system parasitic light and lithography simulation aerial image according in Kirk method, parasitic light at first, the most original definition just comes from Kirk method, therefore lithography simulation and parasitic light evaluation all by strictly according to light in Kirk method in the definition of parasitic light, obtained by the intensity distributions of emulation photoetching aerial image, the ratio of Ji An island aerial image center intensity and the aerial image intensity in Fei Andao district.Judge stray light level and the relation of emulation parasitic light threshold value, if stray light level is greater than emulation parasitic light threshold value, then correspond to situation shown in Fig. 1, illustrates that initially secretly island size d is too small, then by d assignment in d _minand enter simulation calculation next time, if stray light level is less than emulation parasitic light threshold value, then correspond to situation shown in Fig. 2, illustrate that initially secretly island size d is excessive, then by d assignment in d _maxand enter simulation calculation next time; Iterate according to above rule, final d will converge on a certain value d _island, now stray light level will be close until d as far as possible with emulation parasitic light threshold value _island=d, then finishing iteration process, now d _islandit is then the best dark island size for measuring this projection exposure system therefor 6 parasitic light.

Step 3, utilization best dark island size solve spatial frequency corresponding to the expansion intermediate frequency roughness relevant with parasitic light

After asking for best dark island size, in actual measurement parasitic light process, An Dao center is entered in any scattering and luminous energy in field range is parasitic light, it is the luminous energy redistribution caused by other factors except system value aperture and basic aberration 10, this is consistent with parasitic light mentioned in the parasitic light Tolerance assignment in early stage, in extreme ultraviolet etching system, this factor is only considered to the expansion intermediate frequency roughness 11 in mirror finish error, therefore, can according to the dark island size d of the best obtained above _islandthe spatial frequency range of expansion intermediate frequency roughness 11 correspondence is accurately intercepted with system visual field.

As shown in Figure 5, best dark island size d is utilized for ease of illustrating _islandabsorb the principle of the expansion intermediate frequency roughness 11 relevant to parasitic light, projection exposure system therefor 6 is reduced to single mirror 12, the dark island of optimum dimension 3 forms dark island aerial image 13 after single mirror 12, and clear zone 14 forms clear zone aerial image 15 after single mirror 12, because single mirror 12 exists the expansion intermediate frequency roughness 11 relevant with parasitic light, cause incident light 16 that diffraction occurs and produce 0 grade of light 18 and two bundle secondary light 17, ignore the low high diffracting grade of energy level, wherein a branch of secondary light 17 enters into dark island aerial image 13, superpose a bright background, form parasitic light.Dark island aerial image 13 place image planes and clear zone aerial image 15 place image planes are at grade, if the distance between this plane and single mirror 12 bore is l ', optimum dimension dark island 3 place object plane and place, clear zone 14 object plane are at grade, if the distance between this plane and single mirror 12 bore is l, the space periodic of expansion intermediate frequency roughness 11 is Λ, corresponding spatial frequency best dark island is of a size of d _island, system magnification is α, then in conjunction with grating equation Λ sin θ=λ, solve spatial frequency corresponding to the expansion intermediate frequency roughness 11 relevant with parasitic light from two aspects:

From object space picked-up expansion intermediate frequency roughness 11:

$\frac{d_{\mathrm{island}}/2}{l}=\tan \mathrm{\&theta;}$

$\frac{\mathrm{FOV}/2}{l}=\tan {\mathrm{\&theta;}}^{\&prime;}$

Wherein, λ is system exposure wavelength, θ and θ ' is for incident light 16 is in the angle of diffraction expanding different sizes when intermediate frequency roughness 11 occurs diffraction, and θ < θ ', owing to expansion intermediate frequency roughness 11 having the mismachining tolerance of different frequency, therefore its angle of diffraction is not unique yet, but is positioned at the arbitrary value between θ and θ ', can be in the hope of the spatial frequency distribution scope of the expansion intermediate frequency roughness 11 relevant with parasitic light:

$\frac{\sin \mathrm{\&theta;}}{\mathrm{\&lambda;}}<f<\frac{{\sin \mathrm{\&theta;}}^{\&prime;}}{\mathrm{\&lambda;}}$

From image space picked-up expansion intermediate frequency roughness 11:

$\frac{\mathrm{\&alpha;}{d}_{\mathrm{island}}/2}{l^{\&prime;}}=\frac{d_{\mathrm{island}}/2}{l}=\tan \mathrm{\&theta;}$

$\frac{\mathrm{\&alpha;}\&CenterDot;{\mathrm{FOV}}_{\mathrm{image}}/2}{l^{\&prime;}}=\frac{\mathrm{FOV}/2}{l}=\tan {\mathrm{\&theta;}}^{\&prime;}$

As can be seen here, when solving from image space, when solving with object space, conclusion is consistent, it is a kind of convenient to this provide, as in Swarzschild two mirror projection exposure system therefor, due to the equity of light in object space and image space and light reversible, the expansion intermediate frequency roughness 11 of two pieces of catoptrons can be solved respectively from object space and image space.The expansion intermediate frequency roughness spatial frequency range relevant with parasitic light depends on best dark island size, from object space and the image space corresponding spatial frequency range absorbing expansion intermediate frequency roughness 11 relevant with parasitic light in primary and secondary mirror Swarzschild system respectively, make use of equality and the light law of reciprovity in image space in projection exposure system therefor, for the spatial frequency picked-up of Swarzschild system extension intermediate frequency roughness provides conveniently.

Claims (2)

1. mirror finish error analysis method relevant with parasitic light in extreme ultraviolet etching system, is characterized in that, condition and the step of the method are as follows:

Solving of step one, best dark island size

(1) emulation parasitic light threshold value is determined;

(2) minimum value of dark island size is determined according to photoetching principle λ is exposure wavelength, and NA is the image-side numerical aperture of projection exposure system therefor, and the minimum value of this dark island size is the limiting resolution of projection exposure system therefor, namely all can not successful Imagewise exposure under any condition;

(3) the maximal value d of initial dark island size _max=D, the maximal value of this dark island size can the aerial image on Shi An island be by force 0 in any condition Xia Qi center area light;

(4) minimum value of Qu An island size and the mean value of maximal value first time lithography simulation is entered as initial dark island size;

(5) stray light level is obtained according in Kirk method about the definition of system parasitic light and lithography simulation aerial image, judge stray light level and the relation emulating parasitic light threshold value, if stray light level is greater than emulation parasitic light threshold value, illustrates that initially secretly island size d is too small, then by d assignment in d _minand enter simulation calculation next time, if stray light level is less than emulation parasitic light threshold value, illustrates that initially secretly island size d is excessive, then by d assignment in d _maxand enter simulation calculation next time;

(6) repeat step (5), final d will converge on a certain value d _island, until stray light level is equal with emulation parasitic light threshold value, i.e. d=d _island, then d _islandfor the best dark island size;

Step 2, utilization best dark island size solve spatial frequency corresponding to the expansion intermediate frequency roughness relevant with parasitic light

(1) the dark island of optimum dimension forms dark island aerial image after projection exposure system therefor, and clear zone forms clear zone aerial image after projection exposure system therefor, because there is the expansion intermediate frequency roughness relevant with parasitic light in projection exposure system therefor, incident light generation diffraction is caused to produce 0 grade of light and two bundle secondary light, ignore the low high diffracting grade of energy level, wherein a branch of secondary light enters into dark island aerial image, and superposition one bright background, forms parasitic light;

(2) dark island aerial image place image planes and aerial image place, clear zone image planes are at grade, if the distance between this plane and projection exposure system therefor bore is l ', optimum dimension dark island place object plane and place, clear zone object plane are at grade, if the distance between this plane and projection exposure system therefor bore is l, the space periodic of expansion intermediate frequency roughness is Λ, corresponding spatial frequency best dark island is of a size of d _island, system magnification is α, then in conjunction with grating equation Λ sin θ=λ, wherein, λ is system exposure wavelength, solves spatial frequency corresponding to the expansion intermediate frequency roughness relevant with parasitic light from two aspects:

1. from object space picked-up expansion intermediate frequency roughness:

$\frac{d_{\mathrm{island}}/2}{l}=\tan \mathrm{\&theta;}$

$\frac{\mathrm{FOV}/2}{l}=\tan {\mathrm{\&theta;}}^{\&prime;}$

2. from image space picked-up expansion intermediate frequency roughness:

$\frac{{\mathrm{\&alpha;d}}_{\mathrm{island}}/2}{l^{\&prime;}}=\frac{d_{\mathrm{island}}/2}{l}=\tan \mathrm{\&theta;}$

$\frac{\mathrm{\&alpha;}\&CenterDot;{\mathrm{FOV}}_{\mathrm{image}}/2}{l^{\&prime;}}=\frac{\mathrm{FOV}/2}{l}={\tan \mathrm{\&theta;}}^{\&prime;}$

Wherein, θ and θ ' is for incident light is expanding angle of diffraction ultimate value when intermediate frequency roughness occurs diffraction, and θ < θ ', owing to expansion intermediate frequency roughness having the mismachining tolerance of different frequency, therefore actual angle of diffraction is the arbitrary value between θ and θ ', then the spatial frequency distribution scope that relevant with parasitic light expansion intermediate frequency roughness is corresponding is:

$\frac{\sin \mathrm{\&theta;}}{\mathrm{\&lambda;}}<f<\frac{\sin {\mathrm{\&theta;}}^{\&prime;}}{\mathrm{\&lambda;}}$

No matter be solve or solve from image space conclusion from object space unanimously, therefore can solve expansion intermediate frequency roughness respectively from object space and image space.

2. mirror finish error analysis method relevant with parasitic light in extreme ultraviolet etching system according to claim 1, is characterized in that, further comprising the steps of before carry out step one: the theoretical analysis that best dark island size proposes

When a dark island of small size forms the first aerial image after preferred view exposure system, diffraction is there is due to the restriction of system value aperture, thus make the central area of first aerial image on the dark island of small size have light intensity, even use this small size dark island measuring system parasitic light, for preferred view exposure system, its stray light level recorded is not 0 yet, and this contradicts with the preferred view exposure system of hypothesis, illustrates that this small size dark island size Selection is improper;

When a dark island of large scale forms second space picture after more than half real projection exposure system, half real projection exposure system is made up of the expansion intermediate frequency roughness of preferred view exposure system, basic aberration and introducing parasitic light, the central area light intensity of the second space picture on the dark island of large scale is made to be 0, namely according to this large scale dark island measuring system parasitic light, for half real projection exposure system, its stray light level recorded is 0, this contradicts with half real projection exposure system of hypothesis, illustrates that this large scale dark island size Selection is improper;

Based on the consideration of above two kinds of situations, a dark island of optimum dimension is certainly existed between the dark island of small size and the dark island of large scale, when a dark island of optimum dimension forms the 3rd aerial image after projection exposure system therefor, projection exposure system therefor is made up of preferred view exposure system and basic aberration, the central area light intensity of the 3rd aerial image on the dark island of optimum dimension is made to be just 0, this just matches with the projection exposure system therefor of hypothesis, and emulates with utilization the systems compliant carrying out using in the process of parasitic light Tolerance assignment;

Two aspects are measured theoretical in conjunction with etching system emulation and system parasitic light, obtain the conclusion certainly existing the optimum dark island size that measures for parasitic light, instead of choose arbitrarily, make parasitic light become a kind of intrinsic property of system, relevant with lighting condition, optical element level of processing.