CN1209599C - Differential confocal scanning detection method with high spatial resolution - Google Patents

Abstract

The present invention belongs to the technical field of surface fine structure measurement and relates to a differential confocal scanning detection method with high spatial resolution. The method adopts the subtraction between differential confocal microscopic double-reception light path configuration and a double-detector to form a differential confocal signal to measure a workpiece to be measured. Lateral resolution is enhanced by an optical super-resolution confocal microscopic detection method, and longitudinal resolution is enhanced by a differential confocal microscopic detection method so as to achieve the high spatial resolution detection of differential confocal scanning detection. The method can meet the requirements of high spatial resolution, high precision and great measuring ranges and is particularly suitable for measuring surface three-dimensional fine structures, tiny steps, tiny grooves, line width, surface shapes, etc.

Description

Differential confocal scanning detection method with high spatial resolution

Technical field

The invention belongs to microcosmic detection technique field, a kind of method that detects about surface three dimension microtexture, little step, little groove, integrated circuit live width and surface topography particularly is provided.

Background technology:

Involve the development of quantum device to VLSI (very large scale integrated circuit), millimeter along with semiconductor technology, Micrometer-Nanometer Processing Technology has entered the three-dimensional processing technique field of deep-submicron, nanometer, therefore press for research adapt to its demand for development on a large scale, the imaging and the detection technique of high spatial (promptly axially and horizontal) resolution characteristic.

In the detection range of high spatial resolution, contactless scan-probe class detection method can be in leading position undoubtedly.In recent years, although the scan-probe class detection method based on principles such as scanning-tunnelling, atomic force, near field optics has obtained fast development, greatly improved the spatial resolving power of sensor, make it to direct at nanometer scale and (laterally be better than 2nm, axially be better than the 0.1nm nanometer scale), but there is following limitation and not enough in these class methods.Based on all kinds of Scanning Probe Microscopy of technology such as scanning-tunnelling, require during imaging to be controlled in the scope of nanometer scale between probe and the sample; Also require to be controlled in the scope of a wavelength between the probe of Near-field Optical Microscope and the sample and have shortcomings such as optical energy loss is big.Shortcomings such as above-mentioned reason limited the imaging sensing range of this quasi-instrument, restricted it at big kick sample with change application in the big sample of projection, and in addition, also there is the apparatus structure complexity in this kind method, involves great expense, test condition harshness, image taking speed are low.

Because confocal microscope has unique three-dimensional chromatography imaging characteristic, in recent years, the part scholar utilizes it to carry out high resolution imaging and detection both at home and abroad.

For example, the Tan Jiubin of Harbin Institute of Technology, Wang Fusheng, Zhao Weiqian have proposed " differential confocal formula nanoscale optical focus detection method ", make azimuthal resolution reach 2nm.(" scientific and technological scientific seminar collection of thesis is measured in the 3rd both sides of the Straits ". Lanzhou, 2000:59 ~ 63);

Chinese patent " confocal microscope " (application number: 01122439.8, publication number: CN 1395127A) propose interferometric method is introduced in traditional confocal micro imaging system, be used to improve azimuthal resolution;

Chinese patent " double-frequency confocal step height microscope measuring device " (application number: 02120884.0, publication number: CN 1384334A) disclose a kind of double-frequency confocal step and interfered microscopic system;

The C-H.Lee of Taiwan Univ. etc. has proposed non-interference difference confocal microscopy theory (Optics Comm.1997,35:233～237), and it utilizes response curve hypotenuse linearity range to realize that nanoscale detects;

Chinese patent " method of cofocus scanning test material face shape " (application number: 99113699.3, publication number: CN 1274074A) proposition utilizes the face shape on the response characteristic test material surface of confocal microscope;

1998, (Axial superresolution with phase-only pupil filters.Optics Communications.1998 such as American scholar T asso R.M.Sales, 156:227-230) design two districts pure phase position iris filter, improved the azimuthal resolution of optical system.

But above-mentioned achievement only is confined to the improvement and the raising of optical sensor system azimuthal resolution, does not relate to the raising of transverse super-resolution power.

In addition, though improved 1.4 times based on the transverse resolution of confocal microscopy detection system than the transverse resolution of the common microscopic system under the same terms, its transverse resolution is still than low 2 orders of magnitude of azimuthal resolution.Therefore, the raising of confocal microscope system (comprising all light contact pilotage class sensors) transverse resolution has become the key that improves its spatial resolution.

And confocal microscope point illumination and point are surveyed this unique light path arrangement and be convenient to combine with iris filter and realize the characteristic of optical ultra-discrimination providing an otherwise effective technique approach for solving above-mentioned optical ultra-discrimination.

At present, adopting the three-dimensional super-resolution iris filter is the main means that improve optical scanning detection method spatial resolution.

For example, Chinese Academy of Sciences's Shanghai ray machine Deng Xiaoqiang etc. designed a kind of three-dimensional super-resolution iris filter (Chinese laser .2001,28 (5): 459～462), in the hope of reaching the three-dimensional super-resolution effect, its structure is as shown in Figure 1: this three-dimensional super-resolution iris filter is for having certain thickness concentric circles annular substrate, and normalization radius in center circle district is a, and the transmitance in this district is k, middle annular regions cylindrical normalization radius is b, and the outermost circle normalization radius of iris filter is 1.Its ultimate principle is the size by optimal design a, k and b parameter value, reaches the three-dimensional super-resolution effect of iris filter.But this wave filter belongs to annular light leakage-type wave filter, exists optical energy loss big, deficiencies such as transmitance control difficulty.

Spain scholar Manuel Martinez-Corra etc., proposition utilizes amplitude type annular transmission-type wave filter to come the central main lobe of sharpening optical system three-dimensional point spread function, make its main lobe not only in horizontal compression, and axially also obtaining sharpening, and then improve the transverse resolution and the azimuthal resolution (Optics Communications.1999,165:267～278) of optical system.

Calendar year 2001, Chinese Academy of Sciences's Shanghai ray machine Liu Li etc. utilize global optimization approach (CGO algorithm), designed two kinds of different three position facies pattern iris filters (Acta Physica Sinica .2001,50 (1): 48～51) that require at confocal microscope.This method with the axial halfwidth (HWHM) of system strength point spread function or laterally HWHM as the optimization aim function, with three annuluses separately area and phasic difference as bound variable, objective function is carried out optimal design, be used to improve the spatial resolution of optical system.

But above-mentioned three-dimensional iris filter should carry out transverse super-resolution and take into account axial super resolution again, and the three-dimensional super-resolution effect is not remarkable.

Summary of the invention:

The objective of the invention is for overcoming the deficiency of above-mentioned prior art, merge the characteristics separately of optical ultra-discrimination and differential confocal microtechnic, the optical detecting method of the absolute type high spatial resolution of a kind of lateral optical super-resolution, axial nanoscale resolution is provided, realizes the non-contact optical of three-dimensional microstructure, little step, little groove, integrated circuit live width and object surface appearance is detected.

Detection method of the present invention is to adopt the double reception light path arrangement of differential confocal microscopy and double detector to subtract each other detection sample is carried out scanning survey, improve the azimuthal resolution of sample by the differential confocal micro-detecting method, the confocal microscopic method by super-resolution pupil filtering confocal microscopic method or shaping circular light illuminaton improves transverse resolution.

A kind of detection method of the present invention comprises the following step:

(1) incident light is by iris filter 1, polarization spectroscope 2 etc., 4 pairs of samples of measurement object lens 12 through the differential confocal microscopic system carry out scanning survey, and photodetector 10 and photodetector 11 record the intensity curve I that reflection sample 12 convex-concaves change size respectively ₁(v, u ,-u _M) and intensity curve I ₂(v, u ,+u _M), wherein, u is axial normalization optical displacement, v is horizontal normalization optical displacement;

(2) with I ₁(v, u ,-u _M) and I ₂(v, u ,+u _M) differentially subtract each other and carry out normalized, obtain the intensity curve I that corresponding sample 12 convex-concaves change ₃(v, u, u _M);

(3) optimize pin hole 8 and pin hole 9 position u apart from its corresponding condenser focus _M, the azimuthal resolution of raising differential confocal microscopic system;

(4) parameter of iris filters such as optimization amplitude type wave filter, phase-type wave filter, amplitude position phase hybrid filter satisfies G _TWith the designing requirement of S, make the Airy disk main lobe of differential confocal microscopic system obtain sharpening, improve the transverse resolution of differential confocal microscopic system, wherein, G _TThe ratio of response curve halfwidth when iris filter and no pupil wave filter are arranged, S are the ratio of the focus maximum of intensity when iris filter is arranged with no pupil wave filter;

(5) according to I ₃(v, u, u _M) near intensity curve s light intensity magnitude in ab measurement range zero point, reconstruct the surface topography and the micro-scale of sample 12.

Wherein, corresponding pin hole 8 and pin hole 9 are apart from the position u of its corresponding condenser focus _MBy following formula make k (0,0, u _M) maximum comes optimization to determine:

$k(\mathrm{0,0},u_M)=2\·\sin c\left[\frac{u_M}{4}\right]\·\left\{\frac{\left\{\frac{u_M}{4}\right\}\·\cos \left\{\frac{u_M}{4}\right\}-\sin \left\{\frac{u_M}{4}\right\}}{{\left\{\frac{u_M}{4}\right\}}^2}\right\}$

Wherein, sinc (x) is the clock function, that is, and and sinc (x)=sinx/x.

Merge improving the differential confocal scan method of azimuthal resolution and the super-resolution pupil filtering cofocus scanning method of raising transverse resolution, constitute pupil filtering formula differential confocal scanning detection method.The iris filter of employing particular design carries out the mask correction to the pupil function of differential confocal microscopic system, and then changes wavefront, sharpening Airy disk main lobe, finally improves the transverse super-resolution power of differential confocal microscopic system.Iris filter can be phase-type wave filter, an amplitude type wave filter and amplitude position hybrid filter mutually.The raising of azimuthal resolution can realize by the light path arrangement and the differential detection of differential confocal.Like this, just, can reach the purpose that improves spatial resolution.

Another kind of detection method of the present invention comprises the following step:

(1) incident light is by shaping binary optical device 16, polarization spectroscope 2 etc., 4 pairs of samples of measurement object lens 12 through the differential confocal microscopic system carry out scanning survey, and photodetector 10 and photodetector 11 record the intensity curve I that reflection sample 12 convex-concaves change size respectively ₁(v, u ,-u _M) and intensity curve I ₂(v, u ,+u _M), wherein, u is axial normalization optical displacement, v is horizontal normalization optical displacement;

(2) with I ₁(v, u ,-u _M)) and I ₂(v, u ,+u _M) differentially subtract each other and carry out normalized, obtain the intensity curve I that corresponding sample 12 convex-concaves change ₃(v, u, u _M);

(3) optimize pin hole 8 and pin hole 9 position u apart from its corresponding condenser focus _M, the azimuthal resolution of raising differential confocal microscopic system;

(4) the normalization radius ε of preferred shaping circular light, the Airy disk main lobe of sharpening differential confocal microscopic system improves its transverse resolution;

(5) according to I ₃(v, u, u _M) near intensity curve s light intensity magnitude in ab measurement range section zero point, reconstruct the surface topography and the micro-scale of sample 12.

When carrying out transverse super-resolution according to above-mentioned employing shaping circular light illuminaton differential confocal microscopy, the u under the normalization radius ε situation of corresponding shaping circular light _MDetermine by following formula is optimum:

$k(\mathrm{0,0},u_M)=2\×{(1-{\mathrm{\ϵ}}^2)}^2\·\sin c\left[\frac{u_M}{4}(1-{\mathrm{\ϵ}}^2)\right]-\left\{\frac{\left\{\frac{u_M(1-{\mathrm{\ϵ}}^2)}{4}\right\}\·\cos \left\{\frac{u_M(1-{\mathrm{\ϵ}}^2)}{4}\right\}-\sin \left\{\frac{u_M(1-{\mathrm{\ϵ}}^2)}{4}\right\}}{{\left\{\frac{u_M(1-{\mathrm{\ϵ}}^2)}{4}\right\}}^2}\right\}$

u _MFor given ε value down sensitivity k (0,0, u _M) respective value when maximum, when ε=0.25, u _M=± 5.56; ε=0.50 o'clock, u _M=± 6.95; ε=0.75 o'clock, u _M=± 11.91, sinc (x) is the clock function, that is, and and sinc (x)=sinx/x.

And the normalization radius of corresponding shaping circular light is preferred in from 0.1 to 0.9 according to super-resolution requirement ε.

Merge improving the differential confocal scan method of azimuthal resolution and the shaping circular light cofocus scanning method of raising transverse resolution, constitute shaping circular light formula differential confocal scanning detection method.Utilize the ring light irradiation of particular design, can improve the transverse resolution of confocal microscope system, widen the axial linear range ability, suppress the side lobe intensity that microscopic system causes because of out of focus.The raising of azimuthal resolution can realize by the light path arrangement and the differential detection of differential confocal.Like this, just, can reach the purpose that improves spatial resolution.

Detection method of the present invention has following characteristics and good result:

The present invention is owing to merged the axial high-resolution characteristic of pupil filtering formula confocal microscopy transverse super-resolution characteristic and differential confocal light path arrangement method, avoided existing three-dimensional super-resolution iris filter should improve transverse resolution, improve azimuthal resolution again, thereby reduce the shortcoming of the combination property of three-dimensional super-resolution.

Measuring method of the present invention has following characteristics:

1) when improving horizontal resolution characteristic, can expand axially (z to) range ability of measurement characteristics curve s;

2) improved the out of focus characteristic of confocal system;

3) measuring system has absolute tracking zero point and bipolarity tracking characteristics, can realize absolute measurement;

4) differential confocal double reception light path arrangement and double detector subtract each other probe method and can suppress the common-mode noise that ambient condition difference, light source intensity fluctuation, the electric drift of detector etc. cause, significantly improve signal to noise ratio (S/N ratio), sensitivity and the linearity etc. of measuring system.

Description of drawings:

Fig. 1 is the three-dimensional super-resolution iris filter.

Fig. 2 is for adopting the differential confocal micro measurement method synoptic diagram of iris filter technology.

Fig. 3 is normalized curve I ₃(v, u, u _M).

Fig. 4 is the super horizontal resolution light intensity response curve comparison diagram of phase-type pupil filtering.

Fig. 5 is for adopting the synoptic diagram of shaping circular light illuminaton differential confocal micro measurement method.

Fig. 6 is different ε, u _MDown, laterally light intensity response.

Fig. 7 is different ε, u under the out-of-focus appearance _MThe time the response of horizontal light intensity.

Fig. 8 is Grad and detector offset u _MChange curve.

Fig. 9 is the relation curve of ε and axial response signal.

Figure 10 is the relation curve of ε and axial response signal.

Figure 11 is the differential confocal sensing principle based on the binary optical device super-resolution.

Figure 12 is the standard bench height scintigram of afm scan.

Figure 13 is the standard step width scan figure of afm scan.

Figure 14 is a differential confocal sensor step transversal scanning curve.

Figure 15 is the differential confocal sensor step transversal scanning curve behind the employing binary optical transverse super-resolution.

Figure 16 is a step transversal scanning comparison chart.

In the accompanying drawing, 1 iris filter, 2 polarization spectroscopes (PBS), 3 quarter-wave plates, 4 measure object lens, 5 spectroscopes (BS), 6 condensers, 7 condensers, 8 pin holes, 9 pin holes, 10 photodetectors, 11 photodetectors, 12 samples, 13 laser instruments, 14 beam expanders, 15 spatial filtering pin holes, 16 shaping binary optical devices (BOE), 17 adjustable diaphragms, 18 micro-displacement work tables, 19 subtracters, 20 displacement transducers, 21 piezoelectric ceramic actuators (PZT), 22 driving powers, 23 amplification treatment circuits, 24 micro-computer processing systems, the micro-double reception light path of 25 differential confocals.

Embodiment:

As shown in Figure 2, empty frame partly is the micro-double reception light path arrangement 25 of differential confocal, and after incident beam process iris filter 1, polarization spectroscope 2, quarter-wave plate 3 transmissions, measured object lens 4 focus on sample 12 surfaces.Through the measuring light of sample 12 surface reflections, after process quarter-wave plate 3 is polarized spectroscope (PBS) 2 reflections once more, be divided into two-beam by spectroscope (BS) 5 again, and focused on by two identical condensers 6 and 7 respectively.Pin hole 8 and detector 10 place the defocused distance C place of condenser 6 focal planes, and pin hole 9 and detector 11 place the burnt front distance D place of condenser 7 focal planes, and C, D size be M, are u apart from the optics normalization displacement of M correspondence _M

If axially the normalization optical displacement is made as u, laterally the normalization optical displacement is made as v.When sample 12 carried out axial and transversal scanning, detector 10 recorded scanning response curve I ₁(v, u ,-u _M), detector 11 records scanning response curve I ₂(v, u ,+u _M).With response curve I ₁(v, u ,-u _M)) and I ₂(v, u ,+u _M) carry out differentially subtracting each other and carrying out normalized, obtain the normalized response curve I as shown in Figure 3 of this measuring method ₃(v, u, u _M), that is:

$I_3(v,u,u_M)=\frac{I_2(v,u,+u_M)-I_1(v,u,-u_M)}{I_2(v,u,+u_M)+I_1(v,u,-u_M)}---\left(1\right)$

Examine axial resolution characteristic, when sample carried out axial scan, v was that constant is made as c, and following formula can abbreviation be:

$I_3(c,u,u_M){|}_{v=c}=\frac{I_2(c,u,+u_M)-I_1(c,u,-u_M)}{I_2(c,u,+u_M)+I_1(c,u,-u_M)}---\left(2\right)$

As shown in Figure 2, I ₃(c, u, u _M) near the convex-concave of the corresponding sample 12 of intensity curve s light intensity magnitude in ab section zero point changes, and utilizes this surface topography that is worth big or small reconstruct sample and micro-scale.

Its concrete principle is as follows:

When sample 12 is in the some A place of nearly Jiao Qu, the maximal value of the corresponding curve e of the signal that detector 11 detects, a end of the corresponding s curve of differential wave this moment; When sample 12 is in the focal position, the signal that detector 11 detects is in curve e descending branch middle part, and the signal that detector 10 detects is in curve f ascent stage central region, the actual zero point of the corresponding s curve of differential confocal signal; When sample 12 is in the some B place of nearly Jiao Qu, the signal that detector 10 detects is near the peak value of curve f, this moment corresponding s curve the b end; In near the AB zone of sample focus, move the ab section of the corresponding s curve of the signal that detector is surveyed.

From the s curve, as can be seen, compare with the hypotenuse section of single channel confocal microscopy family curve e and f, the characteristic hypotenuse section of s steepening, sensitivity doubles, and promptly azimuthal resolution is improved.

Differential confocal micro measurement method with the iris filter technology improves spatial resolution as one of embodiment of the invention below, and the differential confocal scanning detection method that the present invention is had high spatial resolution further specifies as follows:

Still as shown in Figure 2, empty frame partly is a differential confocal microscopy double reception light path arrangement 25, and iris filter 1 is the ring-like iris filter of N district concentric circles.Under the monochromation illumination condition, have near the distribution of amplitudes of pupil function P (ρ) system focus to be:

$U(v,u)=2{\∫}_0^{1}p\left(\mathrm{\ρ}\right)\·\exp (-\mathrm{ju}{\mathrm{\ρ}}^2/2)\·J_0\left(\mathrm{v\ρ}\right)\mathrm{\ρd\ρ}---\left(3\right)$

ρ-return-change radius, the radial coordinate r on the corresponding receiving plane of v-, the u-correspondence is that the axle of initial point is gone up coordinate Z with the focus, wherein:

$\left\{\begin{array}{c}v=\mathrm{kr}\sin \mathrm{\α}\\ u=\mathrm{kz}{\left(\sin \mathrm{\α}\right)}^2\end{array}\right.---\left(4\right)$

t _jTransmittance function, for the j district _jBe the phase differential in j district, after measuring object lens 4 focusing, the lateral amplitude of vibration response characteristic on the focal plane is:

To the amplitude type iris filter, in the formula (6), _j=C (C is a constant), t _jFor variable (j=1,2,3......, N)

To pure phase bit-type iris filter, in the formula (6), t _j=C (C is a constant), _jFor variable (j=1,2,3......, N)

To amplitude position phase hybrid filter, in the formula (6), t _j, _jBe variable (j=1,2,3......, N)

Analyze with pure phase bit-type iris filter at this, other roughly the same.Consider pure phase bit-type iris filter, then t _j=C (j=1,2,3...... N), makes C=1.

The radius of supposing emergent pupil is R, and incident wavelength is λ, (R _j=a _jR, a ₀=0, a _N=1, ₁=0), for the symmetrical phase-type iris filter of N district circle, focal plane amplitude expression formula is:

$I(v,0)={\left|U(v,0)\right|}^2$

Because $\frac{J_1\left(x\right)}{x}=\frac{1}{2}(1-\frac{x^2}{4\×2}+\frac{x^4}{2\×4^2\×6}+\mathrm{LL})---\left(9\right)$

It is approximate to get two-stage, promptly gets $\frac{J_1\left(x\right)}{x}\≈\frac{1}{2}(1-\frac{x^2}{8}),$ Corresponding light intensity is:

Order $I(v,0)=a-\frac{{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C20041000635900181.png,C20041000635900182.png"\; state="\{\{state\}\}">v</figure-callout>}}^2}{4}b+\frac{{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C20041000635900181.png,C20041000635900182.png"\; state="\{\{state\}\}">v</figure-callout>}}^2}{64}c---\left(11\right)$

Wherein

During no pupil wave filter, $I_0(v,0)=1-\frac{{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C20041000635900181.png,C20041000635900182.png"\; state="\{\{state\}\}">v</figure-callout>}}^2}{4}+\frac{{\mathrm{<figure-callout\; id="4"\; label="v"\; filenames="C20041000635900193.png,C20041000635900202.png"\; state="\{\{state\}\}">v</figure-callout>}}^4}{64}---\left(13\right)$

$\frac{\&PartialD;I}{\&PartialD;v}=0\&DoubleRightArrow;{\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="C20041000635900182.png,C20041000635900191.png"\; state="\{\{state\}\}">v</figure-callout>}}_1=0,{\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C20041000635900181.png,C20041000635900182.png"\; state="\{\{state\}\}">\; <figure-callout\; id="3"\; label="v"\; filenames="C20041000635900182.png,C20041000635900192.png"\; state="\{\{state\}\}">v</figure-callout>\; </figure-callout>}}_{\mathrm{2,3}}=\&PlusMinus;2\sqrt{2}$

The central light strength extreme value is: I ₀(0,0)=1 (14)

When iris filter is arranged, $\frac{{\&PartialD;I}_F}{\&PartialD;v}=0,$ Then $-\frac{b\&CenterDot;\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C20041000635900181.png,C20041000635900182.png"\; state="\{\{state\}\}">v</figure-callout>}}{2}+\frac{{\mathrm{<figure-callout\; id="3"\; label="v"\; filenames="C20041000635900182.png,C20041000635900192.png"\; state="\{\{state\}\}">v</figure-callout>}}^3\&CenterDot;\mathrm{<figure-callout\; id="16"\; label="c"\; filenames="C20041000635900193.png"\; state="\{\{state\}\}">c</figure-callout>}}{16}=0,$ Solve an equation to such an extent that the extreme point coordinate is as follows:

$\left\{\begin{array}{c}{\mathrm{<figure-callout\; id="1"\; label="v\; F"\; filenames="C20041000635900182.png,C20041000635900191.png"\; state="\{\{state\}\}">v</figure-callout>}}_F=0\\ {\mathrm{<figure-callout\; id="2"\; label="v\; F"\; filenames="C20041000635900181.png,C20041000635900182.png"\; state="\{\{state\}\}">v</figure-callout>}}_F=-2\sqrt{2}\\ {\mathrm{<figure-callout\; id="3"\; label="v\; F"\; filenames="C20041000635900182.png,C20041000635900192.png"\; state="\{\{state\}\}">v</figure-callout>}}_F=2\sqrt{2}\sqrt{\frac{b}{c}}\end{array}\sqrt{\frac{b}{c}}\right.---\left(15\right)$

The ratio G of the response curve halfwidth when iris filter and no pupil wave filter are arranged _TFor:

$G_T=\sqrt{\frac{b}{c}}$

The ratio Strehl of the focus maximum of intensity when iris filter and no pupil wave filter are arranged than S is:

Utilize optimal design method, at given G _T, S and ε condition under, determine the N position phase-plate phasic difference in zone separately _jWith normalization radius a _j, optimal conditions is:

Objective function F ( _j, a _j)=G _T-E≤ε, ε=0.02, S 〉=F, 0＜ _j＜2 π, 0＜a _j＜1 a _N=1;

E and the F target for optimizing, optimized Algorithm adopts the Nelder-Mead simplex direct search algorithm on the Matlab6.1 software.

Get N=4, promptly when the phase-type iris filter is 4 district wave filters, choose following three groups of G _TWith S as optimization aim, optimize the iris filter parameter that obtains three groups of correspondences after finding the solution:

1) works as G _T=0.7643, during S=0.25, four corresponding position facies pattern iris filter a ₁=0.1, a ₂=0.2, a ₃=0.5199, a ₄=1, ₁=0, ₂=2.8634rad, ₃=1.5222rad, ₄=5.8372rad;

2) work as G _T=0.8020, during S=0.3, four corresponding position facies pattern iris filter a ₁=0.2, a ₂=0.3058, a ₃=0.5332, a ₄=1, ₁=0, ₂=1.5777rad, ₃=3.0112rad, ₄=5.7177rad;

3) work as G _T=0.8512, during S=0.35, four corresponding position facies pattern iris filter a ₁=0.3, a ₂=0.4, a ₃=0.5804, a ₄=1, ₁=0, ₂=1.6834rad, ₃=3.6583rad, ₄=6.2829rad.

With the super horizontal resolution characteristic curve plotting of above-mentioned three kinds of phase-type iris filters in Fig. 4, therefrom as can be seen: after adding the phase-type iris filter, the transverse response curve obtains sharpening, and G _TBe worth more for a short time, sharpening is obvious more, and it is obvious more that transverse resolution improves.Disadvantage is that secondary lobe strengthens, it is that the S value diminishes that optical energy loss also increases, but secondary lobe can the pin hole by confocal microscope system suppress (this just iris filter combine with confocal microscopy, realize the reason of super diffraction resolved detection truly), optical energy loss can solve by the enlargement factor that increases detection system.

In the iris filter formula differential confocal micro measurement method, the raising of azimuthal resolution can be by optimizing u _MValue realizes, u _MBy formula (19) make k (0,0, u _M) maximum optimizes definite:

$k(\mathrm{0,0},u_M)=2\&CenterDot;\sin c\left[\frac{u_M}{4}\right]\&CenterDot;\left\{\frac{\left\{\frac{u_M}{4}\right\}\&CenterDot;\cos \left\{\frac{u_M}{4}\right\}-\sin \left\{\frac{u_M}{4}\right\}}{{\left\{\frac{u_M}{4}\right\}}^2}\right\}---\left(19\right)$

Improving spatial resolution with shaping circular light formula differential confocal scanning detection method below is two of the embodiment of the invention, and the differential confocal scanning detection method that the present invention is had high spatial resolution further specifies as follows:

As shown in Figure 6, laser instrument 13 sends the laser beam of wavelength X=630nm, expand bundle by beam expander 14 then and be the Gaussian beam of φ 4mm, by spatial filtering pin hole 15, make it to become pointolite, the directional light that expands after restrainting incides shaping binary optical device (BOE) 16, through adjustable diaphragm 17, directional light is shaped as the required annular beam of super-resolution, confocal sensor-based system is thrown light on, annular beam carries out transmission and reflection through polarization spectroscope (PBS) 2, wherein after passing through λ/4 wave plates 3 and microcobjective 4, the s light of PBS reflection focuses on the surface of sample 12, light through sample 12 reflections returns along former road then, once more through becoming p light behind λ/4 wave plates 3 and seeing through polarization spectroscope (PBS) 2, this transmitted light is divided into two equicohesive folded light beam of bundle and transmitted light beams again through semi-transparent semi-reflecting spectroscope 5, and focus on through condenser 7 and condenser 6 respectively, equidistant respectively the burnt preceding and defocused of condenser 6 and condenser 7 that place of pin hole 8 and pin hole 9, and receive its light intensity signal by detector 11 and detector 10 respectively, the light intensity signal that detector 11 and detector 10 are detected subtracts each other through subtracter 19, after amplifying, amplifier 23 just obtains near the focus error signal of differential confocal sensor differential zero point again, the corresponding measured object distance focal point of this signal position size.Range ability for extension sensor, microcobjective 4 is cemented in the object lens Z that is made of piezoelectric ceramic actuator (PZT) 21 and displacement transducer 20 in the spotting scaming system, these object lens Z reaches 350 μ m to the range ability of spotting scaming system, and sweep frequency reaches 150Hz.Micro-computer processing system 24 control piezoelectric ceramic actuators (PZT) 21 are done axial displacement, when sample 12 processes are measured focal plane of lens, detected differential light intensity signal is through near the null value, as the aiming trigger pip, near the signal sum value the null value the when signal that this moment, displacement transducer recorded triggers with aiming can reflect that just the sample axial location changes with it.

The resolution characteristic of the confocal sensor measurement of the absolute tracking mode of present embodiment high spatial resolution draws according to following Theoretical Calculation.

Light intensity response curve function I (v, u, the u of reflective annular confocal microscope _M) be:

$I_1(v,u,u_M)=\left|\right[{\&Integral;}_{{\mathrm{\&epsiv;}}_1}^1{P}_1\left(\mathrm{\&rho;}\right)e^{\left(\mathrm{ju}{\mathrm{\&rho;}}^2\right)/2}{J}_0\left(\mathrm{\&rho;v}\right)\mathrm{\&rho;d\&rho;}]\&CenterDot;[{\&Integral;}_{{\mathrm{\&epsiv;}}_2}^1{P}_1\left(\mathrm{\&rho;}\right)e^{j{\mathrm{\&rho;}}^2(u+u_M)/2}{J}_0\left(\mathrm{\&rho;v}\right)\mathrm{\&rho;d\&rho;}]{|}^2---\left(20\right)$

Wherein:

$u=\frac{8\mathrm{\&pi;}}{\mathrm{\&lambda;}}z{\sin }^2\left(\frac{{\mathrm{\&alpha;}}_0}{2}\right)$ $---\left(21\right)$

Z is for moving axially distance, α ₀Be the numerical aperture of objective angle.

Normalization radius pupil function P (ρ) is,

ε is a laser beam normalization radius, u _MBe the pin hole axial offset.

When utilizing the binary optical device shaping circular light, the light of core is not blocked, but is moved on the outer shroud, and system is the noenergy loss in energy transfer process, supposes that the amplitude on the ring is A, and then beam intensity can be expressed as after the shaping:

$\left\{\begin{array}{cc}I\left(r\right)=0& 0\&le;r\&le;\mathrm{\&epsiv;}\\ I\left(r\right)=A^2& \mathrm{\&epsiv;}\&le;\mathrm{<figure-callout\; id="1"\; label="r"\; filenames="C20041000635900182.png,C20041000635900191.png"\; state="\{\{state\}\}">r</figure-callout>}1\&le;1\end{array}\right.---\left(23\right)$

Get according to law of conservation of energy:

$E=2\mathrm{\&pi;}{\&Integral;}_{\mathrm{\&epsiv;}}^1{\mathrm{<figure-callout\; id="2"\; label="A"\; filenames="C20041000635900181.png,C20041000635900182.png"\; state="\{\{state\}\}">A</figure-callout>}}^2\mathrm{rdr}=2\mathrm{\&pi;}{\&Integral;}_0^1\mathrm{rdr}---\left(24\right)$

Try to achieve: $A=\frac{1}{\sqrt{1-{\mathrm{\&epsiv;}}^2}}---\left(25\right)$

The horizontal resolution characteristic of this measuring method is:

When measured object is in out-of-focus appearance, photodetector axially exists side-play amount, i.e. u _M≠ 0 o'clock, get by formula (20) and formula (25), the transverse intensity distribution characteristic is:

$I_1(v,u,u_M){|}_{u=C}=\left|\right[{\&Integral;}_{\mathrm{\&epsiv;}}^{1}A\&CenterDot;e^{\left(\mathrm{ju}{\mathrm{\&rho;}}^2\right)/2}{J}_0\left(\mathrm{\&rho;v}\right)\mathrm{\&rho;d\&rho;}]\&CenterDot;[{\&Integral;}_{\mathrm{\&epsiv;}}^{1}A\&CenterDot;e^{j{\mathrm{\&rho;}}^2(u+u_M)/2}{J}_0\left(\mathrm{\&rho;v}\right)\mathrm{\&rho;d\&rho;}]{|}^2---\left(26\right)$

It is u=0 that sample 12 is on the focal plane, and laterally resolution characteristic is with the pin hole axial offset u of ring normalization radius ε and optimization in the annular beam _MChanging Pattern as shown in Figure 7.As can see from Figure 7, ε is big more, and the halfwidth of light intensity is just more little to be that transverse resolution is high more, but finds simultaneously, and along with its secondary lobe of increase of ε value also increases, it is unfavorable that this detects imaging, can fortunately be that the pin hole of confocal microscopy can suppress its influence.

Common confocal sensor all is operated under the out of focus situation, supposes u=5, and then horizontal resolution characteristic is with the pin hole axial offset u of ring normalization radius ε and optimization in the annular beam _MChanging Pattern as shown in Figure 8, therefrom can see: the ε value is big more, and the light intensity halfwidth is more little, the Sidelobe Suppression that causes because of out of focus must be obvious more.

The axial resolution characteristic of this measuring method is:

When two point probes depart from picture focal plane+u respectively _MWith-u _M, the normalization radius of annular pupil is ε, is got by formula (25) and formula (26):

$I(0,u,u_M)=|A\&CenterDot;U(0,u,u_M)\&times;A\&CenterDot;U{(0,u,u_M)}^{*}|=\frac{1}{4}(1-{\mathrm{\&epsiv;}}^2)\&CenterDot;\sin c^2\left[\frac{(2u+u_M)}{4}(1-{\mathrm{\&epsiv;}}^2)\right]---\left(27\right)$

Two detectors are differential subtract each other after, focus error signal I ₃(0, u, u _M) be:

$I_3(0,u,u_M)=I(0,u,u_M)-I(0,u,-u_M)---\left(28\right)$

$=(1-{\mathrm{\&epsiv;}}^2)\&CenterDot;\left\{\sin c^2\right[\frac{(2u+u_M)}{4}(1-{\mathrm{\&epsiv;}}^2)]-\sin c^2[\frac{(2u-u_M)}{4}(1-{\mathrm{\&epsiv;}}^2)\left]\right\}$

Corresponding relation between focus error signal and the detection range u as the formula (28), again by formula (21) the relation curve s between focus error signal and the detection range z.

In the ring light differential confocal sensing technology, detector axis is to bias size u _MTo directly influence sensor axis, for determining a suitable u to resolution characteristic _M, make focus error signal I (0, u, u _M) change of sensitivity be the resolving power maximum, with formula (28) to the u differentiate, near the sensitivity expression formula of non-cumulative curve u=0 be:

$k(\mathrm{0,0},u_M)=2\&times;{(1-{\mathrm{\&epsiv;}}^2)}^2\&CenterDot;\sin c\left[\frac{u_M}{4}(1-{\mathrm{\&epsiv;}}^2)\right]\&CenterDot;\left\{\frac{\left\{\frac{u_M(1-{\mathrm{\&epsiv;}}^2)}{4}\right\}\&CenterDot;\cos \left\{\frac{u_M(1-{\mathrm{\&epsiv;}}^2)}{4}\right\}-\sin \left\{\frac{u_M(1-{\mathrm{\&epsiv;}}^2)}{4}\right\}}{{\left\{\frac{u_M(1-{\mathrm{\&epsiv;}}^2)}{4}\right\}}^2}\right\}---\left(29\right)$

Get by formula (21) and formula (29): the sensitivity k of ring light differential confocal probe method response curve, by two pin hole position u _M, the decision of ring light normalization radius ε, numerical aperture of objective.

Fig. 9 has drawn and has worked as u=0, ε=0.00, ε=0.25,0.50 and 0.75 o'clock, gradient k (0, u, u _M) be worth with detector offset u _MThe curve that changes.Can draw by the simulation calculation analysis and by Fig. 7:

Work as u=0, ε=0.00 o'clock is at u _M=± 5.21 places, and k (0, u, u _M) the absolute value maximum, and u _M=-5.21, and k (0, u, u _M)=0.54; u _M=5.21, and k (0, u, u _M)=-0.54;

Work as u=0, ε=0.25 o'clock is at u _M=± 5.56 places, and k (0, u, u _M) the absolute value maximum, and u _M=-5.56, and k (0, u, u _M)=0.4748; u _M=5.56, and k (0, u, u _M)=-0.4748;

Work as u=0, ε=0.50 o'clock is at u _M=± 6.95 places, and k (0, u, u _M) the absolute value maximum, and u _M=-6.95, and k (0, u, u _M)=0.3038; u _M=6.95, and k (0, u, u _M)=-0.3038;

Work as u=0, ε=0.75 o'clock is at u _M=± 11.91 places, and k (0, u, u _M) the absolute value maximum, and u _M=-11.91, and k (0, u, u _M)=0.1034; u _M=11.91, and k (0, u, u _M)=-0.1034.

The axial offset of getting pin hole is above-mentioned corresponding optimum axial dipole field value u _MThe time, the slope absolute value maximum of focusing error curved segments.

When ε=0.50, pin hole axial dipole field u _M=6.95 o'clock, the relation curve that obtains the axial response signal that focus error signal and two detectors survey as shown in figure 10.

When ε=0.75, pin hole axial dipole field u _M=5.56 o'clock, the relation curve that obtains the axial response signal that focus error signal and two detectors survey as shown in figure 11.

ε increases, and the linearity measuring range of ring light differential confocal detection method is widened, and secondary lobe is inhibited, but its resolving power reduces, and when adopting the method to survey, should take all factors into consideration, and takes into account azimuthal resolution, transverse resolution and measurement range, makes combination property reach best.

When adopting shaping circular light illuminaton differential confocal microscopy to carry out transverse super-resolution, adopt preferred ε value to realize.

As shown in Figure 5, empty frame partly is the micro-double reception light path arrangement 25 of differential confocal, and the pin hole 8 under the corresponding ε value and the position of pin hole 9 are u _M, optimize u _MImprove the azimuthal resolution of differential confocal microscopic system, u _MCan determine by formula (29) is optimum.

u _MFor given ε value down sensitivity k (0,0, u _M) respective value when maximum, as when the ε=0.25, u _M=± 5.56; ε=0.50 o'clock, u _M=± 6.95; ε=0.75 o'clock, u _M=± 11.91.

In a single day the pin hole position of optical ultra-discrimination differential confocal system is decided, the actual zero point of response curve s just is determined, the linear zone ab section of getting response curve s is as measuring section, the sensitivity of the corresponding measuring system of its slope size k is by two pin hole positions, iris filter and the decision of numerical aperture of objective size.

The main devices model and the parameter of experimental verification system of differential confocal scanning detection method that the present invention has high spatial resolution is as follows:

As shown in Figure 6, measure object lens 4 in the experiment and preferentially select 10 * 0.25,40 * 0.65 and 60 * 0.85 common flat field achromatic micro objective respectively for use.

Photodetector 10 and the 11 preferential 2001 type photelectric receivers that adopt U.S. NEWFOCUS company to produce, the saturation power scope is 10mW, the maximum adjustable gain is 10 ⁴, the minimal noise equivalent power is 0.25pW/Hz ^1/2, be that the response at 632.8nm place is 0.42A/W at wavelength.

Pin hole 8 and 9 is preferentially selected the PH-10 type pin hole of U.S. NERPORT company for use, and it is made of ultra-thin Mo, and aperture size is 10 μ m, and thickness is 15.24 μ m.

The driver of micro-displacement work table 18 select preferentially that U.S. NEWFOCUS company produces for use on a large scale, high stability Picomotor (micro-displacement driver) driver, the flexible hinge work bench that is equipped with scale down and is 5: 1 is formed nano level fine motion calibration system, and each driving pulse of Picomotor micro-displacement driver can make micro-displacement work table 18 obtain the feeding of 2nm.

Measure the axial tracking location of object lens 4, the microcobjective micropoistioning device that the German PI of preferential employing company produces, it is made of micrometric displacement drive system 21, high accuracy displacement sensor 20 and axial hinge driving mechanism etc., the driving resolving power is 10nm, range 300 μ m, the frequency response after the loading is 100Hz.The inside and outside footpath of ring light after binary optical device 16 shapings is respectively φ 0.856mm and φ 3.484mm, in order to verify the super-resolution performance under the different ε value situations, uses adjustable diaphragm 17 to change the external diameter of annular pupil.Use the diaphragm of φ 3.5mm, φ 1.7mm and φ 1.1mm to change the external diameter of annular beam respectively at this, make its corresponding normalization radius be respectively ε ≈ 0.25, ε=0.5 and ε ≈ 0.78.

Super-resolution performance The tested results based on the confocal sensor surveying unit of the absolute tracking mode of the high spatial resolution of the inventive method is:

The resolution characteristic of system can be examined by the standard step that the measurement U.S. Dimension3100 of DI company type atomic force microscope is worn.Figure 12 and Figure 13 are the scintigram through the standard step of Dimension3100 type afm scan.The bench height that Figure 12 has provided two identification points (triangle) vertical direction correspondence approximates 118.23nm, and the distance of the step skip zone of two identification point horizontal direction correspondences is 0.1367 μ m.The step width that Figure 13 has provided two identification points (triangle) horizontal direction correspondence is 1.836 μ m.

Measured object 12 is selected this standard step for use, microcobjective is selected 60 * 0.85 object lens for use, step is placed on the objective table, adjust step vertically by micro-adjusting mechanism, the light contact pilotage is focused on the ledge surface, then, move step along the horizontal direction vertical with the light contact pilotage, the resolving power of micro displacement workbench is 2nm, the about 13 μ m of moving range, with the amount of movement of HP5528A two-frequency laser interferometer detection step, its resolving power is 0.01 μ m, and drive system can be the amount of feeding fine motion step of 0.01 μ m with resolving power.

ε among Figure 14=0 when not carrying out super-resolution with binary optical device, the step transversal scanning curve that records, the distance of the step skip zone of two identification points (arrow) horizontal direction correspondence is 0.403 μ m, its value comprises the skip zone width of step self.

ε among Figure 15=0.5 is when carrying out super-resolution with binary optical device, the step transversal scanning curve that records, the distance of the corresponding step of the distance skip zone of the identification point horizontal direction shown in two arrows, its size is 0.268 μ m, if consider that again the gradient of step self is 0.1367 μ m (transverse resolution that comprises atomic force microscope self), then transverse resolution is better than 0.2 μ m after the super-resolution.

For ease of comparing, it is shift value that magnitude of voltage is demarcated, step scanning curve when the step scanning curve that does not adopt binary optical device shown in Figure 14 and the employing binary optical device shown in Figure 15 are carried out super-resolution is plotted among Figure 16, the skip zone slope variation big for carry out the displacement curve of super-resolution with binary optical device.

Below in conjunction with the accompanying drawings the specific embodiment of the present invention and test effect are described; but these explanations can not be understood that to have limited scope of the present invention; protection scope of the present invention is limited by the claims of enclosing, and any change of carrying out on claim of the present invention basis all is protection scope of the present invention.

Claims (5)

1, the differential confocal scanning detection method that has high spatial resolution, two light path arrangement (25) and the double detectors accepted that adopt the differential confocal microscopy subtract each other detection sample (12) are carried out scanning survey, it is characterized in that:

＜1〉incident light is by iris filter (1), polarization spectroscope (2) quarter-wave plate (3), measurement object lens (4) through the differential confocal microscopic system carry out scanning survey to sample (12), and photodetector (10) and photodetector (11) record the intensity curve I that reflection sample (12) convex-concave changes size respectively ₁(v, u ,-u _M) and intensity curve I ₂(v, u ,+u _M), wherein, u is axial normalization optical displacement, v is horizontal normalization optical displacement;

＜2〉with I ₁(v, u ,-u _M) and I ₂(v, u ,+u _M) differentially subtract each other and carry out normalized, obtain the intensity curve I that corresponding sample (12) convex-concave changes ₃(v, u, u _M);

＜3〉optimize pin hole (8) and pin hole (9) position u apart from its corresponding condenser focus _M, the azimuthal resolution of raising differential confocal microscopic system;

＜4〉parameter of optimization amplitude type wave filter, phase-type wave filter, amplitude position these iris filters of phase hybrid filter satisfies G _TWith the designing requirement of S, make the Airy disk main lobe of differential confocal microscopic system obtain sharpening, improve the transverse resolution of differential confocal microscopic system, wherein, G _TThe ratio of response curve halfwidth when iris filter and no pupil wave filter are arranged, S are the ratio of the focus maximum of intensity when iris filter is arranged with no pupil wave filter;

＜5〉according to I ₃(v, u, u _M) near intensity curve s light intensity magnitude in ab measurement range section zero point, reconstruct the surface topography and the micro-scale of sample (12).

2, differential confocal scanning detection method according to claim 1, when it is characterized in that adopting the iris filter technology to carry out transverse super-resolution, corresponding pin hole (8) and pin hole (9) are apart from the position u of its corresponding condenser focus _MBy following formula make k (0,0, u _M) maximum comes optimization to determine:

$k(\mathrm{0,0},u_M)=2\&CenterDot;\sin c\left[\frac{u_M}{4}\right]\&CenterDot;\left\{\frac{\left\{\frac{u_M}{4}\right\}\&CenterDot;\cos \left\{\frac{u_M}{4}\right\}-\sin \left\{\frac{u_M}{4}\right\}}{{\left\{\frac{u_M}{4}\right\}}^2}\right\}$

Wherein, sinc (x) is the clock function, that is, and and sinc (x)=sinx/x.

3, the differential confocal scanning detection method that has high spatial resolution, two light path arrangement (25) and the double detectors accepted that adopt the differential confocal microscopy subtract each other detection sample (12) are carried out scanning survey, it is characterized in that:

＜1〉with incident light by shaping binary optical device (16), polarization spectroscope (2) quarter-wave plate (3), measurement object lens (4) through the differential confocal microscopic system carry out scanning survey to sample (12), and photodetector (10) and photodetector (11) record the intensity curve I that reflection sample (12) convex-concave changes size respectively ₁(v, u ,-u _M) and intensity curve I ₂(v, u ,+u _M), wherein, u is axial normalization optical displacement, v is horizontal normalization optical displacement;

＜2〉with I ₁(v, u ,-u _M) and I ₂(v, u ,+u _M) differentially subtract each other and carry out normalized, obtain the intensity curve I that corresponding sample (12) convex-concave changes ₃(v, u, u _M);

＜3〉optimize pin hole (8) and pin hole (9) position u apart from its corresponding condenser focus _M, the azimuthal resolution of raising confocal microscope system;

＜4〉the normalization radius ε of preferred shaping circular light, the Airy disk main lobe of sharpening differential confocal microscopic system, the transverse resolution of raising confocal microscope system;

＜5〉according to I ₃(v, u, u _M) near intensity curve s light intensity magnitude in ab measurement range section zero point, reconstruct the surface topography and the micro-scale of sample (12).

4, differential confocal scanning detection method according to claim 3, when it is characterized in that adopting shaping circular light illuminaton differential confocal microscopy to carry out transverse super-resolution, the u under the normalization radius ε situation of corresponding shaping circular light _MBy following formula make k (0,0, u _M) maximum next optimum definite:

$k(\mathrm{0,0},u_M)=2\&times;{(1-{\mathrm{\&epsiv;}}^2)}^2\&CenterDot;\sin c\left[\frac{u_M}{4}(1-{\mathrm{\&epsiv;}}^2)\right]\&CenterDot;\left\{\frac{\left\{\frac{u_M(1-{\mathrm{\&epsiv;}}^2)}{4}\right\}\&CenterDot;\cos \left\{\frac{u_M(1-{\mathrm{\&epsiv;}}^2)}{4}\right\}-\sin \left\{\frac{u_M(1-{\mathrm{\&epsiv;}}^2)}{4}\right\}}{{\left\{\frac{u_M(1-{\mathrm{\&epsiv;}}^2)}{4}\right\}}^2}\right\}$

u _MFor given ε value down sensitivity k (0,0, u _M) respective value when maximum, when ε=0.25, u _M=± 5.56; ε=0.50 o'clock, u _M=± 6.95; ε=0.75 o'clock, u _M=± 11.91, sinc (x) is the clock function, that is, and and sinc (x)=sinx/x.

5, differential confocal scanning detection method according to claim 3, when it is characterized in that adopting shaping circular light illuminaton differential confocal microscopy to carry out transverse super-resolution, it is preferred in from 0.1 to 0.9 that the normalization radius of corresponding shaping circular light requires ε according to super-resolution.

Images