CN105372948A - Rapid modeling based wave aberration detection method for large-numerical aperture photoetching projection lens - Google Patents

Abstract

A rapid modeling based wave aberration detection method for a large-numerical aperture photoetching projection lens comprises two stages of rapid modeling and aberration extraction; at the stage of rapid modeling, parameters such as polarization state of light, polarization aberration of the projection lens and numerical aperture are firstly set, a space image is simulated by an unary linear sampling method, principal component analysis and multiple linear regression analysis are carried out on the simulation space image to obtain corresponding principle component and a regression matrix, and a detection model matched with a large-numerical aperture photoetching machine is built; and at the stage of aberration extraction, an actually-measured space image is acquired, principal component fitting is carried out on the actually-measured space image to obtain a principle component coefficient, and least square fitting is carried out on the principal component coefficient by using the regression matrix to obtain Zernike aberration of the actually-measured space image. By the rapid modeling based wave aberration detection method, the rapid modeling and the high-precision detection on the Zernike aberration Z5-Z37 of the large-numerical aperture photoetching projection lens are achieved.

Description

Based on the large-numerical aperture wave aberration of photoetching projection objective detection method of rapid modeling

Technical field

The present invention relates to photoetching projection objective lens, particularly a kind of large-numerical aperture wave aberration of photoetching projection objective detection method based on rapid modeling.

Background technology

Litho machine is one of nucleus equipment of great scale integrated circuit manufacture.Projection objective is one of most important subsystem of litho machine.The wave aberration of projection objective is the principal element affecting litho machine alignment precision and imaging resolution.Along with photoetching technique is developed to immersion from dry type, the tolerance for aberration of projection lens of lithography machine becomes more and more harsh, to wave aberration detect speed and accuracy requirement also more and more higher.In order to meet the requirements such as litho machine alignment precision and imaging resolution, a kind of quick, high-precision large-numerical aperture wave aberration of photoetching projection objective detection technique of research and development is significant.

The wave aberration of photoetching projection objective detection technique measured based on aerial image is a common class technology, has that detection speed is fast, cost is low, can detect the advantage of wave aberration of photoetching projection objective in real time.2015, the people such as all bohrs propose a kind of large-numerical aperture wave aberration of photo-etching machine projection objective detection method (see in first technology 1, all bohrs, Li Sikun, Wang Xiangchao, Yan Guanyong, Shen Lina, Wang Lei, " a kind of large-numerical aperture wave aberration of photo-etching machine projection objective detection method ", number of patent application: 201510166998.7, publication number: 104777718A).The method adopts Box-BehnkenDesign statistical sampling methods, uses polarized illumination and vector imaging model, establishes the detection model matched with large-numerical aperture litho machine, achieve the measurement (Z of 33 rank Zernike polynomials fitting ₅~ Z ₃₇), but it is longer to set up detection model required time, is unfavorable for detecting wave aberration of photoetching projection objective fast.

Summary of the invention

The object of the present invention is to provide a kind of large-numerical aperture wave aberration of photoetching projection objective detection method based on rapid modeling, the wave aberration of large-numerical aperture photoetching projection objective lens can be detected fast, accurately.

Technical solution of the present invention is as follows:

Based on a large-numerical aperture wave aberration of photoetching projection objective detection method for rapid modeling, the measuring system that the method adopts comprise for generation of the light source of laser beam, illuminator, for bearing test mask and have precise positioning ability mask platform, for the certification mark on mask graph being imaged onto projection objective system on silicon chip, can silicon chip be carried and the work stage with 3-D scanning ability and precise positioning ability, the data handling machine that is arranged on the aerial image sensor in this work stage and is connected with aerial image sensor.

Described light source can be traditional lighting, ring illumination, two pole illuminations, quadrupole illuminating and free lighting source, and the partial coherence factor of conventional illumination sources is σ; The partial coherence factor of ring illumination light source is [σ _out, σ _in], σ _outrepresent the externally coherent factor, σ _inrepresent inner coherence factor; The partial coherence factor of two pole illuminations is [σ _out, σ _in], σ _outrepresent the externally coherent factor, σ _inrepresent inner coherence factor, pole subtended angle is θ; The partial coherence factor of quadrupole illuminating is [σ _out, σ _in], σ _outrepresent the externally coherent factor, σ _inrepresent inner coherence factor, pole subtended angle is θ.

Described illuminator is for adjusting light distribution and the polarization state of the illumination light field of described light source generation.

Described certification mark is made up of the isolated sky that 6 have different directions orientation, and 6 different direction orientations are respectively 0 °, 30 °, 45 °, 90 °, 120 °, 135 °.

The method comprises rapid modeling and aberration extracts two benches.

The rapid modeling stage comprises the following steps:

1) unitary line sampling method is adopted to set 33 rank Zernike polynomials fitting Z ₅~ Z ₃₇combination ZU, and the Polarization aberration PT of random setting one group of large-numerical aperture projection lens of lithography machine, the sampling matrix D formula of 33 × 67 is as follows:

2) photoetching simulation parameter is selected: the lighting system of illuminator and partial coherence factor thereof, lighting system is polarization illumination, the polarization state of illumination light can be complete polarization, partial polarization and completely unpolarized, photo-etching machine exposal wavelength X, the numerical aperture NA of projection objective, the span of setting NA is NA >=0.93;

3) in mask platform, place test mask, the test badge on this test mask is isolated idle pattern;

4) aerial image acquisition range: X-direction acquisition range is [-L, L], the span of setting L is 300nm≤L≤3000nm, and Z-direction acquisition range is [-F, F], and the span of setting F is aerial image collection is counted: X-direction collection is counted as M, and the span of setting M is M>=20, and Z-direction collection is counted as N, and the span of setting N is N>=13; Above-mentioned parameter and Zernike polynomials fitting combination ZU are inputted computing machine, uses the vector imaging formula that formula is 2. shown, adopt lithography simulation software to emulate, obtain simulation space image set and close AIU.

$\begin{array}{c}\hat{I}(x,y)=n_{image}\underset{-\mathrm{\∞}}{\overset{+\mathrm{\∞}}{\∫}}...\∫J(f,g)H(f+f^{\′},g+g^{\′})H^{*}(f+f^{\′\′},g+g^{\′\′})\\ O(f^{\′},g^{\′})M_0(f+f^{\′},g+g^{\′})E_0\·O^{*}(f^{\′\′},g^{\′\′})M_0^{*}(f+f^{\′\′},g+g^{\′\′})E_0^{*}\\ e^{-i2\mathrm{\π}\[(f^{\′}-f^{\′\′})x+(g^{\′}-g^{\′\′})y\]}{\mathrm{dfdgdf}}^{\′}{\mathrm{dg}}^{\′}{\mathrm{df}}^{\′\′}{\mathrm{dg}}^{\′\′}\end{array}$ ②

Wherein, n _imagefor the refractive index of image space, J (f, g) for normalized efficient light sources intensity distributions, H (f, g) be pupil function, the diffraction spectra that O (f, g) is mask, M ₀(f, g) is the transmission matrix of 3 × 2, E ₀for the Jones vector of incident light, ^*represent conjugate transpose, x and y, f and g are respectively normalized image coordinates, pupil plane coordinate, and normalization formula is as follows:

$\begin{array}{cc}x=\frac{x}{\mathrm{\λ}/NA},& y=\frac{y}{\mathrm{\λ}/NA},\\ f=\frac{f}{NA/\mathrm{\λ}},& g=\frac{g}{NA/\mathrm{\λ}},\end{array}$ ③

Wherein, NA is the numerical aperture of projection objective, and λ is photo-etching machine exposal wavelength, x and y, f and g are respectively image coordinates, pupil plane coordinate.

5) close AIU to simulation space image set and carry out principal component analysis (PCA), obtain the major component of simulation space picture and corresponding major component coefficient, formula is as follows:

AIU＝PC·V④

Wherein, PC is the major component that simulation space image set closes, and V is corresponding major component coefficient.

6) described major component coefficient V and described Zernike polynomials fitting are combined ZU as given data, adopt least square fitting method to calculate linear regression matrix RM, formula is as follows:

V＝ZU·RM⑤

The aberration stage of extracting comprises the following steps:

1) carry out optimum configurations to litho machine to be detected, parameter is identical with the modelling phase;

2) litho machine is started, the illumination light that light source sends obtains the lighting system corresponding to the modelling phase after illuminator adjustment, be irradiated to the test mask in mask platform, the aerial image that the multi-direction test badge utilizing aerial image sensor measurement to converge through projection objective is corresponding, obtain surveying aerial image, and input described computer stored.

3) utilize computing machine to carry out major component matching to actual measurement aerial image, obtain the major component coefficient of surveying aerial image, then carry out matching with described linear regression matrix RM according to least square method, obtain the Zernike polynomials fitting of surveyed photoetching projection objective lens.

With compared with first technology, the present invention has the following advantages:

The present invention, by unitary line sampling mode, reduces hits, simplifies modeling process; And adopting polarized illumination mode and vector imaging model, the accurate characterization aerial image of large-numerical aperture litho machine, finally achieves large-numerical aperture photoetching projection objective lens Zernike polynomials fitting Z ₅~ Z ₃₇rapid modeling and high precision test.

Accompanying drawing explanation

Fig. 1 detection system structural drawing of the present invention.

Fig. 2 lighting system schematic diagram of the present invention, wherein, (a) is the structure of lighting system, and (b) is the polarization state of illumination light.

Fig. 3 mask mark structure of the present invention schematic diagram.

The large-numerical aperture wave aberration of photoetching projection objective precision figure that Fig. 4 uses the present invention's measurement to obtain.

Embodiment

Below in conjunction with embodiment and accompanying drawing, the invention will be further described, but should not limit the scope of the invention with this embodiment.

Fig. 1 is the detection system structural representation that the present invention adopts.Produce the light source 1 of laser beam, illuminator 2, for bearing test mask 3 and have precise positioning ability mask platform 4, for the certification mark 5 on mask graph being imaged onto projection objective system 6 on silicon chip, can silicon chip be carried and the work stage 7 with 3-D scanning ability and precise positioning ability, the data handling machine 9 that is arranged on the aerial image sensor 8 in this work stage 7 and is connected with aerial image sensor 8.

The method comprises rapid modeling and aberration extracts two benches.

The rapid modeling stage comprises the following steps:

1) setting of unitary line sampling method is adopted 33 rank Zernike polynomials fitting Z of amplitude range ₅~ Z ₃₇combination ZU and the Polarization aberration PT of random setting one group of large-numerical aperture photoetching projection objective lens, the sampling matrix D formula of 33 × 67 is as follows:

2) selected lithography simulation parameter: the lighting system of illuminator chooses ring illumination, and its partial coherence factor is [σ _out, σ _in]=[0.9,0.7], the polarization state of illumination light chooses linearly polarized light, and the direction of vibration of this linearly polarized light light vector is parallel with X-direction, as shown in Figure 2, and photo-etching machine exposal wavelength X=193nm, the numerical aperture NA=1.35 of projection objective;

3) in mask platform, place test mask, the test badge on this test mask is isolated idle pattern, and this combination has 6 isolated skies with different directions orientation, and described 6 isolated empty direction orientations are respectively 0 °, 30 °, 45 °, 90 °, 120 °, 135 °, as shown in Figure 3;

4) aerial image acquisition range: X-direction acquisition range is [-900nm, 900nm], Z-direction acquisition range is [-2000,2000]; Aerial image collection is counted: it is 61 that X-direction collection is counted, and it is 57 that Z-direction collection is counted; Above-mentioned parameter design and Zernike polynomials fitting combination ZU are inputted computing machine, uses the vector imaging formula that formula is 2. shown, adopt lithography simulation software to emulate, obtain simulation space image set and close AIU.

$\begin{array}{c}\hat{I}(x,y)=n_{image}\underset{-\mathrm{\∞}}{\overset{+\mathrm{\∞}}{\∫}}...\∫J(f,g)H(f+f^{\′},g+g^{\′})H^{*}(f+f^{\′\′},g+g^{\′\′})\\ O(f^{\′},g^{\′})M_0(f+f^{\′},g+g^{\′})E_0\·O^{*}(f^{\′\′},g^{\′\′})M_0^{*}(f+f^{\′\′},g+g^{\′\′})E_0^{*}\\ e^{-i2\mathrm{\π}\[(f^{\′}-f^{\′\′})x+(g^{\′}-g^{\′\′})y\]}{\mathrm{dfdgdf}}^{\′}{\mathrm{dg}}^{\′}{\mathrm{df}}^{\′\′}{\mathrm{dg}}^{\′\′}\end{array}$ ②

Wherein, n _imagefor the refractive index of image space, J (f, g) for normalized efficient light sources intensity distributions, H (f, g) be pupil function, the diffraction spectra that O (f, g) is mask, M ₀(f, g) is the transmission matrix of 3 × 2, E ₀for the Jones vector of incident light, ^*represent conjugate transpose, x and y, f and g are respectively normalized image coordinates, pupil plane coordinate, and normalization formula is as follows:

$\begin{array}{cc}x=\frac{x}{\mathrm{\λ}/NA},& y=\frac{y}{\mathrm{\λ}/NA};\\ f=\frac{f}{NA/\mathrm{\λ}},& g=\frac{g}{NA/\mathrm{\λ}};\end{array}$ ③

Wherein, NA is the numerical aperture of projection objective, and λ is photo-etching machine exposal wavelength, x and y, f and g are respectively image coordinates, pupil plane coordinate.

5) close AIU to simulation space image set and carry out principal component analysis (PCA), obtain the major component of simulation space picture and corresponding major component coefficient, formula is as follows:

AIU＝PC·V④

Wherein, PC is the major component that simulation space image set closes, and V is corresponding major component coefficient.

6) described major component coefficient V and described Zernike polynomials fitting are combined ZU as given data, adopt least square fitting method to calculate linear regression matrix RM, formula is as follows:

V＝ZU·RM⑤

The aberration stage of extracting comprises the following steps:

1) carry out optimum configurations to litho machine to be detected, parameter is identical with the modelling phase;

2) litho machine is started, the illumination light that light source sends obtains the lighting system corresponding to the modelling phase after illuminator adjustment, be irradiated to the test mask in mask platform, the aerial image that the multi-direction test badge utilizing aerial image sensor measurement to converge through projection objective is corresponding, obtain surveying aerial image, and input described computer stored.

3) computing machine is utilized to carry out major component matching to actual measurement aerial image, obtain the major component coefficient of surveying aerial image, then carry out matching with described linear regression matrix RM according to least square method, obtain the Zernike polynomials fitting of surveyed photoetching projection objective lens, as shown in Figure 4.The rapid modeling part time used is 10min, and the average error of the Zernike polynomials fitting that detection obtains and standard deviation are all at below 0.07nm.

Relative in first technology, the present invention, by unitary line sampling mode, reduces hits, simplifies modeling process, highly shortened the modeling time; And adopting polarized illumination mode and vector imaging model, the accurate characterization aerial image of large-numerical aperture litho machine, finally achieves large-numerical aperture photoetching projection objective lens Zernike polynomials fitting Z ₅~ Z ₃₇rapid modeling and high precision test.

The illustrative embodiments that above embodiment is only used to principle of the present invention is described and adopts, but the present invention is not limited thereto.For those skilled in the art; without departing from the spirit and substance in the present invention; can make various modification and improvement, therefore all equivalent technical schemes also belong to category of the present invention, and scope of patent protection of the present invention should be defined by the claims.

Claims (5)

1. the large-numerical aperture wave aberration of photoetching projection objective detection method based on rapid modeling, the measuring system that the method adopts comprises the light source (1) for generation of laser beam, illuminator (2), a mask platform (4) of precise positioning ability is had for bearing test mask (3), for the certification mark (5) on mask graph being imaged onto the projection objective system (6) on silicon chip, can silicon chip be carried and there is the work stage (7) of 3-D scanning ability and precise positioning ability, the data handling machine (9) being arranged on the aerial image sensor (8) in this work stage (7) and being connected with aerial image sensor (8),

It is characterized in that, the method comprises rapid modeling and aberration extracts two benches;

The described rapid modeling stage comprises the following steps:

1) unitary line sampling method is adopted to set 33 rank Zernike polynomials fitting Z ₅~ Z ₃₇combination ZU, and random setting one group of Polarization aberration PT, the sampling matrix D formula of 33 × 67 is as follows:

2) photoetching simulation parameter is selected: the lighting system of illuminator, i.e. the numerical aperture NA of the partial coherence factor of polarization illumination, light source, photo-etching machine exposal wavelength X and projection objective;

3) in mask platform, place test mask (3), the test badge on this test mask is isolated idle pattern;

4) installation space is counted as acquisition range and collection: X-direction acquisition range is counted as M for [-L, L] and collection, and Z-direction acquisition range is that [-F, F] and collection are counted as N;

5) by above-mentioned steps 1)-step 4) parameters input computing machine, adopt lithography simulation software to emulate, obtain simulation space image set and close AIU;

6) close AIU to simulation space image set and carry out principal component analysis (PCA), obtain the major component of simulation space picture and corresponding major component coefficient, formula is as follows:

AIU＝PC·V④

Wherein, PC is the major component that simulation space image set closes, and V is corresponding major component coefficient;

7) described major component coefficient V and described Zernike polynomials fitting are combined ZU as given data, adopt least square fitting method to calculate linear regression matrix RM, formula is as follows:

V＝ZU·RM⑤

The described aberration stage of extracting comprises the following steps:

1) carry out optimum configurations to litho machine to be detected, parameter is identical with the modelling phase;

2) litho machine is started, the illumination light that light source sends obtains the lighting system corresponding to the modelling phase after illuminator adjustment, be irradiated to the test mask in mask platform, the aerial image that the multi-direction test badge utilizing aerial image sensor measurement to converge through projection objective is corresponding, obtain surveying aerial image, and input described computer stored;

3) utilize computing machine to carry out major component matching to actual measurement aerial image, obtain the major component coefficient of surveying aerial image, then carry out matching with described linear regression matrix RM according to least square method, obtain the Zernike polynomials fitting of surveyed photoetching projection objective lens.

2. the large-numerical aperture wave aberration of photoetching projection objective detection method based on rapid modeling according to claim 1, it is characterized in that, described light source is traditional lighting, ring illumination, two pole illuminations, quadrupole illuminating or free lighting source, and the partial coherence factor of conventional illumination sources is σ; The partial coherence factor of ring illumination light source is [σ _out, σ _in], σ _outrepresent the externally coherent factor, σ _inrepresent inner coherence factor; The partial coherence factor of two pole illuminations is [σ _out, σ _in], σ _outrepresent the externally coherent factor, σ _inrepresent inner coherence factor, pole subtended angle is θ; The partial coherence factor of quadrupole illuminating is [σ _out, σ _in], σ _outrepresent the externally coherent factor, σ _inrepresent inner coherence factor, pole subtended angle is θ;

Described illuminator is for adjusting light distribution and the polarization state of the illumination light field of described light source generation;

Described certification mark is made up of the isolated sky that 6 have different directions orientation, and 6 different direction orientations are respectively 0 °, 30 °, 45 °, 90 °, 120 ° and 135 °.

3. the large-numerical aperture wave aberration of photoetching projection objective detection method based on rapid modeling according to claim 1, is characterized in that, the polarization state of described polarization illumination is complete polarization, partial polarization or completely unpolarized.

4. the large-numerical aperture wave aberration of photoetching projection objective detection method based on rapid modeling according to claim 1, is characterized in that, the span of described X-direction acquisition range L is 300nm≤L≤3000nm; The span of Z-direction acquisition range F is 2000nm≤F≤6000nm; The count span of M of X-direction collection is M >=20, and the span that Z-direction gathers points N is N >=13.

5. the large-numerical aperture wave aberration of photoetching projection objective detection method based on rapid modeling according to claim 1, is characterized in that, numerical aperture NA >=0.93 of described projection objective.