CN110260788A - Optical micro/nano measuring device, the method for extracting structure micro-nano dimension information to be measured - Google Patents

Abstract

The present invention provides a kind of optical micro/nano measuring devices and the method for automatically extracting object to be measured dimension information.The illumination part and imaging moiety of the measuring device are made of multiple 4f systems, and different illumination numerical apertures, the reflective Kohler illumination mode of polarization state can be generated according to object to be measured type and sensitivity requirement and obtains the picture of required enlargement ratio.The device realizes the purpose of logical burnt scanning using a piezoelectric positioner with the objective scan of itself, and swash width can be controlled accurately, and be easy to be directly mounted on line of production technology, can satisfy the requirement of on-line measurement.The size extraction method contains the noise cancelling alorithm based on Fourier transformation.The Fourier spectrum and phase of out-of-focus image are emulated by calculating the practical out-of-focus image of each width and corresponding ideal, and by the former frequency spectrum and the latter's phase combination, its inverse Fourier transform is calculated again, eliminates the optical noise generated in measurement process by external interference and mechanical noise to realize.

Description

Optical micro-nano measurement device, method for extracting micro-nano size information of structures to be measured

technical field

The invention relates to a through-focus scanning optical micro-nano measurement device based on Fourier transform correction, a method for extracting micro-nano size information of a structure to be measured, and a recording medium. It specifically relates to an optical nanometer measurement device capable of correcting optical and mechanical noise and obtaining higher measurement accuracy results and a method for extracting micronano size information of structures to be measured, belonging to the field of optical micronano measurement and process control in the semiconductor industry.

Background technique

Driven by Moore's Law, the size of semiconductors continues to decrease, and its structure gradually develops from a 2-dimensional plane to a 3-dimensional three-dimensional direction. However, the traditional micro-nano measurement method has gradually exposed its limitations. For example, the measurement efficiency of atomic force microscope is too low, and scanning electron microscope can only provide 2-dimensional width information but cannot accurately provide 3-dimensional height information. Compared with traditional micro-nano measurement methods, optical micro-nano measurement has a series of unique advantages, such as low cost, high measurement efficiency, strong operability, and non-destructive. The optical micro-nano measurement method based on through-focus scanning means that for a set of structural parameters x=[x ₁ ,...,x _n ] ^T , the corresponding optical signal set f under the continuous change of the focus position of the objective lens is calculated in advance by means of simulation. (x). Then use the measuring device to collect the optical signal y(x ₀ ) corresponding to the structural parameter x ₀ to be measured, and calculate the root mean square error of the two:

χ(x, x ₀ )=||f(x)-y(x ₀ )||

Finally, combined with the idea of least squares, the measurement results closest to the real structural parameters are extracted from the root mean square error which satisfies:

where Ω is the structural parameter field.

Under practical application conditions, the measurement results of optical micro-nano measurements based on through-focus scanning are highly sensitive to some factors, such as the lateral mechanical vibration of the measurement device and the angular non-uniformity of the actual illumination light. The above factors will introduce mechanical and optical noise, which will reduce the accuracy of the measurement results. Therefore, the Fourier transform algorithm is used to correct the optical signal set f(x) to obtain measurement results that eliminate the interference of mechanical noise and optical noise, and improve the accuracy of measurement results.

Contents of the invention

The technical problem to be solved by the present invention is to provide a through-focus scanning optical micro-nano measurement device that can quickly correct mechanical noise and optical noise, and is helpful for on-line and real-time measurement of the micro-nano size structure of semiconductor components, and is used to extract the A method for measuring the micro-nano size information of structures.

The optical micro-nano measurement device involved in the present invention includes: a reflective Kohler illumination and imaging part; and an objective lens positioning part, wherein the reflective Kohler illumination and imaging part includes: a light source; Converging illuminating light; the small hole is placed in front of the first objective lens; the first lens is placed in front of the small hole to form a collimated optical path; the second lens is placed in front of the first lens for Forming a 4f system with the first lens; a polarizer, placed in the collimated optical path; a dichroic prism, placed behind the polarizer; and a third lens, placed behind the second lens; a fourth lens, Placed behind the third lens, used to form a 4f system with the third lens; a plane mirror; a second objective lens, used to generate reflective Kohler illumination and imaging; and imaging CCD, the objective lens positioning part includes: piezoelectric positioning a device, connected to the second objective lens, used to make the second objective lens scan up and down along the optical axis to achieve the purpose of through-focus scanning; and a control unit, controlling the piezoelectric positioner.

In the measurement system of the present invention, preferably, the light source, the first objective lens and the small hole form a light source system, and the first objective lens converges the illumination light emitted from the light source to the center of the small hole , the aperture then changes the focused illumination light into divergent illumination light.

In the measurement device of the present invention, it is preferred that the first lens and the second lens form a 4f system, and the incident focal plane of the 4f system formed by the first lens and the second lens is the same as the center of the small hole. Where the plane coincides, the exit focal plane of the 4f system composed of the first lens and the second lens coincides with the conjugate rear focal plane of the second objective lens, and different incident numerical apertures are obtained by adjusting the diameter of the small hole Reflective Kohler illumination pattern.

In the measuring device according to the present invention, preferably, the polarizer is placed between the first lens and the second lens, and for different types of samples to be measured, by adjusting the direction of the optical axis of the polarizer to obtain the optimum Good measurement sensitivity.

In the measuring device of the present invention, preferably, the control unit controls the piezoelectric positioner so that the second objective lens scans up and down along the optical axis to realize through-focus scanning without any influence on the stage, and at the same time The mechanical noise generated by the mechanical displacement of the measurement device itself and the optical noise generated by defocusing in the required scanning stroke are small, and have little impact on the accuracy of the measurement results.

The method for extracting the micro-nano size information of the structure to be tested includes: a simulated image simulation step for generating a sequence of simulated out-of-focus images forming a search library; an actual image collection step for collecting a sequence of defocused images of the structure to be measured ; The actual image correction step is used to remove the optical noise and mechanical noise introduced by external interference in the actual image; the step of establishing the through-focus scanning image; and the step of comparing the through-focus scanning image and extracting the information to be measured.

The method for extracting the micro-nano size information of the structure to be measured according to the present invention preferably uses the full vector calculation method in the analog image simulation step to first decompose the Koehler illumination light into a series of plane wave components, and then calculate the corresponding plane wave components of each plane wave component. Scattering near-field distribution, and then use the Abbe imaging principle to calculate the far-field transformation and image plane distribution of each near-field distribution, and finally perform superposition and synthesis to obtain the final ideal analog image.

In the method for extracting the micro-nano size information of the structure to be measured according to the present invention, preferably in the actual image acquisition step, by controlling the piezoelectric positioner, the objective lens can be corrected back to the initial position after each scanning stroke without manual operation. Adjustment, to meet the needs of rapid repeated measurement.

In the method for extracting the micro-nano size information of the structure to be measured according to the present invention, preferably in the actual image correction step, the idea of Fourier transform is used to separate the spectrum and phase of the noise-containing actual image and the ideal simulated image respectively, and the The phase of the real image with noise and the frequency spectrum of the ideal analog image are combined to eliminate the optical noise and mechanical noise introduced by external interference in the original real image.

In the computer-readable storage medium of the present invention, executable instructions are stored thereon, and when the instructions are executed by one or more processors, the one or more processors can execute the method described in claims 6-9. A method for extracting micro-nano size information of the structure to be measured.

Compared with the prior art, the present invention has the following advantages:

(1) The present invention is simple in structure, easy to operate, eliminates image noise from the software point of view, and does not require high hardware level;

(2) The present invention uses the scanning method of the objective lens to replace the stage scanning method of the traditional through-focus scanning system to realize the purpose of through-focus scanning, so that the system is easy to be installed in semiconductor production and processing equipment;

(3) The measurement result of the present invention has high precision and fast measurement speed, and is suitable for on-line measurement and mass production applications.

Description of drawings

Fig. 1 is a measurement system model diagram of the present invention;

Fig. 2 is illumination 4f light path figure of the present invention;

Fig. 3 is imaging 4f optical path figure of the present invention;

Fig. 4 is a block diagram of reflective Kohler illumination and imaging part of the present invention;

Fig. 5 is a block diagram of the objective lens positioning part of the present invention:

Fig. 6 is the overall flowchart of the method for extracting the micro-nano size information of the structure to be tested according to the present invention;

Fig. 7 is the flowchart of the analog image emulation step of the present invention;

Fig. 8 is a flowchart of the actual image acquisition steps of the present invention;

Fig. 9 is a flowchart of the actual image correction steps of the present invention;

Fig. 10 is a flow chart of the steps of establishing a through-focus scanning image of the present invention;

Fig. 11 is a flow chart of the steps of comparison of through-focus scanning images and extraction of information to be tested in the present invention.

Detailed ways

Fig. 1 is the model schematic diagram of the measurement system 10 of the embodiment of the present invention, and this system is made up of following three parts: reflective Kohler illumination part, imaging part and objective lens positioning part, wherein reflective Kohler illumination part comprises: LED light source 11, A first objective lens 21 , a small hole 22 , a first lens 23 , a polarizing plate 24 , a non-polarizing beam splitter prism 25 , a second lens 26 , a flat mirror 27 , a second objective lens 28 , and an object stage 12 . The imaging part includes: a second objective lens 28 , a piezoelectric positioner 29 , a plane mirror 27 , a second lens 26 , a non-polarizing beam splitter prism 25 , a third lens 31 , a fourth lens 32 , and a CCD camera 33 . The objective lens positioning part includes a piezoelectric positioner 29 and its control unit 14 . The second objective lens 28 is controlled by a piezoelectric positioner 29 . When the measurement starts, the control unit 14 sends an instruction to 29, and 29 controls the second objective lens 28 to move downward for a certain stroke from the nominal focus position, and then conducts a through-focus scan from bottom to top to obtain a series of defocused images without adjusting the object. At the same time, the mechanical noise generated by the mechanical displacement of the device itself and the optical noise generated by defocus in the required scanning stroke are relatively small, which has little influence on the accuracy of the measurement results.

Fig. 2 is a schematic diagram of the principle of the reflective Kohler illumination part in the embodiment of the present invention. A first objective lens 21 is placed in front of the LED light source 11 for converging incident light. A small hole 22 is placed at the point of convergence. A first lens 23 with a focal length of 100nm is placed in front of 22, its focal point coincides with the small hole 22, and the divergent beam emitted from 22 is converged into a collimated beam. A polarizing plate 24 is placed in the collimating light path for generating different polarized illumination lights, so as to select appropriate illumination states for different types of targets to be measured. A non-polarizing beam splitter prism 25 is placed behind 24 for turning the light path. Another second lens 26 with a focal length of 100 nm is placed behind 15 , and the second lens 26 and 23 together image the divergent light beam emitted from 22 on the focal plane of a second objective lens 28 . Adjust the position of the second objective lens 28 so that its conjugate back focal plane coincides with the focal plane of the second lens 26, and now the divergent spherical wave sent by each image point on the focal plane of the second lens 26 becomes Plane waves of different propagation directions, thus forming a reflected Koehler illumination pattern on the object plane. The first lens and the second lens constitute the relay lens group of the reflective Kohler illumination part, and form the incident focal plane coincident with the plane where the small hole is located, and the exit focal plane is conjugated with the second objective lens The 4f system with coincident focal planes can easily obtain different illumination numerical apertures by adjusting the size of the small hole diameter.

Fig. 3 is a schematic diagram of the principle of the imaging part of the embodiment of the present invention. The second objective lens 28 collects the scattered light from the sample 13 and forms an image together with the second lens 26 . The third lens 31 is placed behind the second lens 26 , and its focal plane coincides with the focal plane of the second lens 26 to converge the scattered light emitted from the second lens 26 into collimated light. The fourth lens 32 is placed behind the third lens 31 to converge the collimated light onto the photosensitive plane of the CCD camera 33 . By selecting the focal length of the fourth lens 32 and adjusting the distance between the fourth lens 32 and the third lens 31 , images with different magnifications can be obtained.

Fig. 4 is the block diagram of reflective Kohler illumination and imaging part according to the embodiment of the present invention, wherein aforesaid second objective lens 28, aforesaid flat reflector 27, aforesaid second lens 26, aforesaid non-polarizing beam splitter prism 25 are illumination and imaging The common module of the building plays the role of transmitting illumination light and scattering light at the same time.

Fig. 5 is a diagram of an objective lens positioning module according to an embodiment of the present invention. The control unit 14 sends a movement command to the piezoelectric positioner 29 to control the precise movement of the second objective lens 28 . While the second objective lens 28 is moving, it feeds back the number of image acquisitions to the control unit 14 , and when the number reaches a predetermined number, the control unit 14 issues an instruction to stop the movement.

Fig. 6 is an overall flow chart of the method for extracting the micro-nano size information of the structure to be measured according to the present invention. The method for extracting the micro-nano size information of the structure to be measured in the present invention includes: an analog image simulation step 1, which is used to pre-simulate the images of the simulated samples corresponding to n different sizes that make up the search library before measuring the actual target to be measured. n ideal simulated defocused images, the simulation method of each ideal simulated defocused image is based on the idea of full vector calculation, that is, first use the time domain finite difference method to calculate the scattering near-field distribution and its far-field transformation, and then use The Abbe imaging principle superimposes and synthesizes all the far-field components passing through the imaging aperture on the image plane to obtain the light intensity distribution on the image plane; the actual image acquisition step 2 is used to quickly acquire the defocused image of the structure to be measured after starting the measurement Sequence, wherein the control part outputs a control signal to make the piezoelectric positioner drive the objective lens to scan along the optical axis. After the objective lens steps a certain distance, the CCD camera immediately shoots an actual defocused image of the target to be measured; the actual image correction step 3, It is used to use the Fourier transform correction algorithm to remove the optical noise and mechanical noise introduced by external interference in the actual image after the completion of step 2, that is, firstly, the amplitude spectrum of each actual defocused image and the ideal simulated defocused image and the phase spectrum are separated by Fourier transform, and then the amplitude spectrum of the actual defocused image is combined with the phase spectrum of the ideal simulated defocused image, and finally it is inversely Fourier transformed to achieve the actual defocused image about the ideal The purpose of simulated out-of-focus image correction; the through-focus scan image establishment step 4 is used to create the actual through-focus scan image and the simulated through-focus scan image from the actual out-of-focus image and the simulated out-of-focus image after step 1 and step 3 are completed, respectively Image search library, in which for each actual out-of-focus image and simulated out-of-focus image, firstly select a sub-area containing the part of interest in the target to be tested, then extract the average intensity of the sub-area, and finally average these Intensity is stacked according to the focus position in step 1 and step 2 and interpolated and smoothed to output the through-focus scan image; compared with the through-focus scan image and the step 5 of extracting the information to be tested is used to extract the to-be-measured information after step 4 is completed. Measure the size information of the target, that is, based on the idea of the least square method, quantitatively calculate the difference between the actual through-focus image and each ideal simulated through-focus image in the search library, and the best matching result is the difference between the actual through-focus image The simulation model parameters corresponding to the smallest ideal simulation through-focus image are taken as the corresponding parameters of the target to be measured.

Fig. 7 is a flow chart of the simulation image simulation steps. Step 61 is to input the structural parameter x of the model, the optical parameters of the measurement system 10: wavelength λ, illumination numerical aperture INA and collection numerical aperture CNA before running the simulation program, and set the calculation domain size L _x and _Ly of FDTD. In the Koehler illumination mode, every point in the field of view receives the same illumination light cone, and the spatial angle of the illumination light cone is related to INA. Therefore, in step 62, the illumination light cone is decomposed into a series of plane wave components with the same amplitude and different propagation directions, and then in step 63, the scattering near-field distribution corresponding to each plane wave component affected by the sample to be tested is calculated by using the time domain finite difference method. After obtaining the scattered near-field distribution, step 64 transforms the near-field distribution into the far-field distribution in k-space according to the principle of grating diffraction, that is, the Fourier transform of the scattered field. The above steps all act independently on the components in the X, Y, and Z directions of the light vector, so they belong to the method of full vector calculation. According to Abbe's imaging principle, the image plane distribution is actually the superposition result of all scattered Fourier components that can pass through the image square aperture defined by the CNA, so step 65 outputs the simulated separation by superimposing the Fourier components of the scattered field out of focus image.

Fig. 8 is a schematic flow chart of the actual image acquisition steps. Before acquisition, set the camera exposure time, input the number of out-of-focus images 2n+1, through-focus scan step size k nm and specify the image storage path. After the collection starts, the piezoelectric positioner 29 is controlled so that the piezoelectric positioner 29 firstly controls the second objective lens 28 to move downward by n×knm, and then the control unit 14 performs an iterative calculation. The camera 33 exposes once every iteration, and then the second objective lens 28 moves up by k nm, and judges whether the number of captured images meets the requirement. When the number of captured images meets the requirements, the control unit 14 sends an instruction to the piezoelectric positioner 29 to control the second objective lens 28 to return to the initial position. By controlling the piezoelectric positioner, the objective lens can be corrected back to the initial position after each scanning stroke without manual adjustment, which meets the needs of rapid repeated measurement.

Fig. 9 is a flowchart of the actual image correction steps. Optical noise will cause the pattern of the through-focus scanning image to tilt, and mechanical noise will cause the pattern of the through-focus scanning image to fluctuate. It can be seen that both of them are caused by displacement-related effects. Step 71 applies Fourier transform to each defocused image to obtain the frequency spectrum and phase of the actual defocused image and the ideal simulated defocused image respectively. Then step 72 multiplies the phase of the simulated defocused image by the frequency spectrum of the actual defocused image to obtain the Fourier transform of the corrected image, and finally applies the inverse Fourier transform to the result in step 73, then the corrected one can be obtained out of focus image.

Figure 10 is a flow chart of the steps for creating a through-focus scanning image. The through-focus scanning image is actually a two-dimensional cross-sectional view of a three-dimensional scattering intensity distribution data set formed by stacking out-of-focus image sequences, and its X-axis represents each The real spatial position of the part, the Y axis represents the focal position corresponding to each defocused image, and the gray value represents the normalized light intensity. Among them, step 81 uses the smooth background defocus image to perform normalization operation on the defocus image containing the target, and the gray value of the background area of the normalized defocus image obtained tends to 0, which is beneficial to eliminate the intensity of the illumination light in the acquisition process. The impact of fluctuations. The intensity profile extracted in step 82 is the average of the intensity of the sub-region, so any dimensional changes smaller than the sub-region cannot be resolved. Steps 83 and 84 respectively perform interpolation and smoothing filtering on the stacked rough-pass focal scan images to eliminate discontinuities caused by stacking.

Figure 11 is a flow chart of the steps of through-focus scanning image comparison and information extraction. Extracting the size information of the target to be measured is actually a process of solving an inverse problem. First, the through-focus image of the target to be measured is compared with the The obtained series of simulated through-focus images are matched to obtain a series of quantitative matching indicators, and then the fitting functions of these matching indicators are established, and the size information of the target to be measured is extracted by finding the model size corresponding to the minimum value of the fitting function. Wherein step 91 calculates the difference between the through-focus image of the target to be measured and the simulated through-focus image, and uses it as a quantitative matching index. Step 92 selects a quadratic function or a spline function as the fitting function according to the distribution of the matching index. Step 93 calculates the minimum value of the fitting function. When multiple local extreme points appear, further processing is required to find an optimal solution.

Some block diagrams and/or flowcharts are shown in the figures. It will be understood that some or combinations of blocks in the block diagrams and/or flowcharts can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, so that these instructions, when executed by the processor, can be created to implement the functions illustrated in these block diagrams and/or flowcharts /operated device.

Accordingly, the techniques of the present disclosure may be implemented in the form of hardware and/or software (including firmware, microcode, etc.). Additionally, the technology of the present disclosure may take the form of a computer program product on a computer-readable medium having instructions stored thereon for use by or in connection with an instruction execution system (e.g., one or more processors) . In the context of the present disclosure, a computer-readable medium is any medium that can contain, store, convey, propagate or transport instructions. For example, a computer readable medium may include, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, device, or propagation medium. Specific examples of computer-readable media include: magnetic storage, such as magnetic tape or hard disk (HDD); optical storage, such as compact disc (CD-ROM); memory, such as random access memory (RAM) or flash memory; and/or wired /wireless communication link".

While example embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that changes may be made to these example embodiments without departing from the scope and spirit of the invention as defined in the claims and their equivalents. Variations in various forms and details.

Claims (10)

1. a kind of optical micro/nano measuring device, it is characterised in that:

It include: reflective Kohler illumination and imaging moiety；With

Object lens position portion,

Wherein reflective Kohler illumination and imaging moiety include:

Light source；

First object lens, are placed in front of light source, for assembling illumination light；

Aperture is placed in front of first object lens；

First lens are placed in front of the aperture, are used to form collimated light path；

Second lens are placed in front of first lens, are used for and the first lens forming 4f system；

Polarizing film is placed in collimated light path；

Amici prism is placed in after the polarizing film；With

The third lens are placed in after second lens；

4th lens are placed in the third lens rear, for forming 4f system with the third lens；

Plane mirror；

Second object lens, for generating reflective Kohler illumination and imaging；With

CCD is imaged,

Object lens position portion includes: piezoelectric positioner, is connect with second object lens, for making second object lens along optical axis The purpose of logical burnt scanning is realized in lower scanning；With

Control unit controls the piezoelectric positioner.

2. measuring system according to claim 1, it is characterised in that: the light source, first object lens and the aperture A light-source system is constituted, illumination light sent from the light source is converged to the small hole center, institute by first object lens Stating aperture becomes divergent illumination light for convergence illumination light again.

3. measuring device according to claim 1, it is characterised in that:

First lens and second lens constitute 4f system, the 4f of first lens and second lens composition The incident focal plane of system is overlapped with plane where the small hole center, the 4f of first lens and second lens composition The outgoing focal plane of system is overlapped with the conjugation back focal plane of second object lens, by adjusting hole diameter size, to obtain The reflective Kohler illumination mode of different incident numerical apertures.

4. measuring device according to claim 1, it is characterised in that:

The polarizing film is placed between first lens and second lens, for different types of sample to be tested, is led to The light passing axis direction of adjustment polarizing film is crossed, to obtain optimal measurement sensitivity.

5. measuring device according to claim 1, it is characterised in that:

The control unit controls the piezoelectric positioner, so that second object lens are scanned up and down along optical axis and realized logical Coke scanning does not generate any influence, while mechanical noise caused by the mechanical displacement of the measuring device itself to objective table And optical noise caused by required swash width internal cause defocus is smaller, influences on the precision of measurement result smaller.

6. a kind of method for extracting structure micro-nano dimension information to be measured, it is characterised in that:

Include:

Analog image simulation process, for generating the emulation out-of-focus image sequence of composition search library；

Actual image acquisition step, for acquiring the out-of-focus image sequence to geodesic structure；

Real image aligning step, for removing in real image by the introduced optical noise of external interference and mechanical noise；

Logical coke scan image establishment step；With

Logical coke scan image compares and information extracting step to be measured.

7. the method according to claim 6 for extracting structure micro-nano dimension information to be measured, it is characterised in that:

It is first a series of planes by Kohler illumination photodegradation using full vector calculation method in analog image simulation process Then wave component calculates the corresponding NEAR FIELD SCATTERING distribution of each plane wave component, recycles Abbe theory of image formation to calculate each The Far-Zone Field Transformation of a near field distribution and as plane distribution is finally overlapped synthesis and obtains final ideal analog image.

8. the method according to claim 6 for extracting structure micro-nano dimension information to be measured, it is characterised in that:

In actual image acquisition step,

By control piezoelectric positioner, make object lens after each swash width by feedback compensation return to initial position without It must manually adjust, meet quick duplicate measurements needs.

9. the method according to claim 6 for extracting structure micro-nano dimension information to be measured, it is characterised in that:

In real image aligning step, using the thought of Fourier transformation, it respectively is isolated by Noise real image and ideal The frequency spectrum and phase of analog image, and by the spectral combination of the phase of Noise real image and ideal analog image, to eliminate By the introduced optical noise of external interference and mechanical noise in former real image.

10. a kind of computer readable storage medium is stored thereon with executable instruction, described instruction is in one or more processors When execution, one or more of processor perform claims can be made to require described in 6~9 for extracting structure micro-nano to be measured The method of dimension information.