阅读说明：本技术 基于针尖盲重构的光栅原子力显微镜图像重构方法 (Grating atomic force microscope image reconstruction method based on needle point blind reconstruction ) 是由 韩国强 李怀东 邹宇 吕路遥 牛弋翔 于 2019-12-05 设计创作，主要内容包括：本发明涉及一种基于针尖盲重构的光栅原子力显微镜图像重构方法,包括以下步骤：步骤S1：利用OP玻璃表面或TGZ3光栅作为针尖表征器件来获取AFM针尖形貌的信息；步骤S2:根据得到的AFM针尖形貌信息,利用MATLAB进行尖端盲重构,得到重构的AFM针尖形状；步骤S3:根据AFM测量光栅过程中探针与样品表面的几何关系构建线宽测量模型；步骤S4:获得TGZ3光栅样品实测的表面形貌；步骤S5:根据重构的AFM针尖形状和线宽测量模型,对实测的TGZ3光栅表面形貌进行矫正,得到真实的光栅表面结构。本发明在不改变探针硬件设施的条件下,减小由探头驱动器PZT的非线性,系统的温度漂移和探头磨损引起的测量误差,有效地减少和消除AFM图像的失真,从而获得更精确的AFM图像。(The invention relates to a grating atomic force microscope image reconstruction method based on needle point blind reconstruction, which comprises the following steps: step S1: obtaining the information of AFM (atomic force microscope) needle point morphology by using an OP glass surface or a TGZ3 grating as a needle point characterization device; step S2, carrying out tip blind reconstruction by using MATLAB according to the obtained AFM tip morphology information to obtain a reconstructed AFM tip shape; s3, constructing a line width measurement model according to the geometric relationship between the probe and the sample surface in the grating measurement process by AFM; s4, obtaining the actually measured surface topography of the TGZ3 grating sample; and S5, correcting the actually measured surface morphology of the TGZ3 grating according to the reconstructed AFM pinpoint shape and the line width measurement model to obtain a real grating surface structure. Under the condition of not changing the hardware facility of the probe, the invention reduces the measurement error caused by the nonlinearity of the probe driver PZT, the temperature drift of the system and the abrasion of the probe, and effectively reduces and eliminates the distortion of the AFM image, thereby obtaining a more accurate AFM image.)

1. A grating atomic force microscope image reconstruction method based on needle point blind reconstruction is characterized by comprising the following steps:

step S1: using an OP glass surface or a TGZ3 grating as a tip characterization device to acquire AFM tip morphology information;

step S2, carrying out tip blind reconstruction by using MATLAB according to the obtained AFM tip morphology information to obtain a reconstructed AFM tip shape;

s3, constructing a line width measurement model according to the geometric relationship between the probe and the sample surface in the grating measurement process by AFM;

s4, obtaining the actually measured surface topography of the TGZ3 grating sample;

and S5, correcting the actually measured surface morphology of the TGZ3 grating according to the reconstructed AFM pinpoint shape and the line width measurement model to obtain a real grating surface structure.

2. The method for reconstructing the grating atomic force microscope image based on the blind reconstruction of the needle point as claimed in claim 1, wherein the blind tip measurement model of the AFM is approximately expressed by a mathematical morphology development expression:

wherein, I is a point set of the scanned image surface, S is a point set describing sample similarity, and T is a point set describing tip state reflection;

the AFM tip shape estimation function is as follows:

wherein T describes the shape of the needle tip, and d represents the position variation of T in I; the shape of the tip can be expressed by the following formula:

wherein T is_i+1Is the ith result T_i' the (i +1) th iteration result, T_i' (x) is T_iAt x, bringing I into contact with the points contained in I.

3. The method for reconstructing the grating atomic force microscope image based on the blind needle point reconstruction as claimed in claim 2, wherein the step S2 specifically comprises:

step S21, obtaining an AFM image from the profile of the sample scanned by the given tip, including three local maximum points, respectively labeled 1,2 and 3;

step S22: independently taking two needle points obtained in image areas near the maximum values 1 and 2, then calculating the current local needle point shape through a formula (3), and then taking the intersection of two derived needle points (1 and 2);

step S23: the reconstructed AFM tip shape is obtained by merging the derived tip 3 with the previous results.

4. The method for reconstructing the grating atomic force microscope image based on the blind needle point reconstruction as claimed in claim 1, wherein the step S3 specifically comprises:

step S31, to obtain a line width measurement model, firstly defining the shape of the AFM tip as a cone with two unequal semi-cone angles α and β, wherein the side generatrix of the cone is the radius R of the sphere, a is the total cone vertex angle of the outer side generatrix, and delta is the installation angle of the AFM probe;

step S32, introducing the position state parameter of the AFM tip moving along the scanning direction in the sample structure, wherein phi represents the tip scanning inclination angle when the AFM tip scans on the grating;

and step S33, obtaining a geometric relation between the line width profile and the tip through a profile image with a sharp edge structure obtained through AFM signals, wherein the formula is as follows:

a＝α+β (8)

L₍₀₎considering that the probe head affects the measured scanning line width in order to have a height of 0; l is_(h)Considering the probe head influence on the measured scanning line width for the height h;

because the edge radius R of the sharp edge is less than 1nm and far less than the tip radius R (R > R), neglecting, when AFM is used to measure the side wall of the line width structure, the direction of the structure line is not perpendicular to the scanning direction, the transverse distance of the line width structure in the image does not represent the real line width, and the real line width W is the real line width_LineCan be expressed as equation (9):

W_Line＝L₀cosθ (9)

where θ represents an angle between the scanning direction and the structure line.

And finally, a line width measurement model can be constructed according to comparison between the real geometric shape of the cross section of the tested line width structure and the shape obtained from the AFM and the relationship between the line width and the probe tip particles.

Technical Field

The invention relates to the field of atomic force microscopes, in particular to a grating atomic force microscope image reconstruction method based on needle point blind reconstruction.

Background

An Atomic Force Microscope (AFM) is a local probe scanning technology with nanoscale in-plane resolution and sub-angstrom vertical resolution, and makes up for the defects of the traditional microscopic measurement technology. In contrast to atomic force microscopy, SEM and TEM can only provide two-dimensional images, while AFM can provide precise topographical features and three-dimensional (3D) metrology at nanometer and even atomic resolution without destroying the sample. This is because AFM does not require any special treatment of the sample (e.g., copper or carbon plating) which can cause irreversible damage to the sample. In addition, the force applied by the probe tip to the sample surface can cause damage to the sample, especially soft surface samples such as biological cells. Therefore, it is important to improve the measurement speed and reduce the interaction between the tip and the sample while ensuring the imaging quality. At present, two methods for realizing AFM high-efficiency measurement are mainly available. First solution in order to make the AFM tip move faster on the sample, improve the imaging quality, and realize a high-speed atomic force microscope, various new physical elements such as a small cantilever, a micro-resonator, a new actuator, etc. have been designed. However, the complex hardware design and modification of standard AFMs will incur expensive hardware costs. A second solution is to employ new controllers and algorithms such as feedforward and feedback control, a combination of robust controllers and iterative control methods, etc.

Due to measurement errors in the AFM, it is necessary to reconstruct AFM images to obtain high-precision scan images. The measurement error of AFM is mainly caused by the nonlinearity of PZT probe driver, the temperature drift of system sample stage or the vibration of probe, the shape of needle tip, etc. The non-linearity problem of the probe actuator PZT can be solved by model-based compensation and closed-loop control based on a position sensor. The temperature drift of the system can be dynamically compensated by local scanning real-time feedback. Errors caused by AFM tips are most difficult to eliminate because the morphology and state of the probe changes as the scanning process changes. Therefore, how to eliminate the error caused by real-time characterization of the probe is the key to solving the adverse effect of the measurement. Generally, methods of determining the shape of an AFM tip can be roughly classified into two types, i.e., in-situ methods and ex-situ methods. For the ex-situ method, a method of observing a probe such as SEM and TEM with an electron microscope having a high spatial resolution is commonly used, but it is difficult to obtain accurate three-dimensional information of the probe only by the electron microscope. In the in-situ method, the in-situ characterization of the probe can be roughly divided into two categories, one is to calibrate the shape of the needle by scanning a well-defined reference sample surface or a needle characterizer, and the reference sample comprises a calibrated micro/nano structure, a standard calibration grating and the like. However, the morphology of these nanomaterials or standards does not strictly meet the assumptions and is difficult to manufacture, which makes it difficult to accurately estimate the characterization of the tip. Another class is the use of uncalibrated samples in conjunction with a dead-end reconstruction (BTR) algorithm to obtain the optimal upper boundary of the tip shape without knowledge of the tip characteristics and state. This method simplifies the calibration process of the tip morphology, but the accuracy of tip characterization depends on whether the reference sample has multiple nanostructures. Moreover, if a higher tip is to be characterized, the reference sample must have a higher feature-ratio of the nanostructures. More importantly, after characterization of the probe tip, an accurate AFM image can be reconstructed using the etching operation. However, the accuracy of AFM images reconstructed by etching methods is too low to meet the expected image accuracy requirements, and the erosion and expansion operations cannot be used to eliminate measurement errors due to interaction between the probe and the sample. There is a need to develop an effective image reconstruction method for different samples to eliminate image distortion caused by the morphology and state of the probe tip.

Disclosure of Invention

In view of the above, the present invention provides a method for reconstructing an atomic force microscope image of a grating based on a blind reconstruction of a tip, which can effectively reduce and eliminate distortion of an AFM image by reducing a possible deviation between the AFM image and a shape of an actual sample due to a shape and a position state of the AFM tip without changing a probe hardware facility, thereby obtaining a more accurate AFM image.

In order to achieve the purpose, the invention adopts the following technical scheme:

a grating atomic force microscope image reconstruction method based on needle point blind reconstruction comprises the following steps:

step S1: obtaining the information of AFM (atomic force microscope) needle point morphology by using an OP glass surface or a TGZ3 grating as a needle point characterization device;

step S2, carrying out tip blind reconstruction by using MATLAB according to the obtained AFM tip morphology information to obtain a reconstructed AFM tip shape;

s3, constructing a line width measurement model according to the geometric relationship between the probe and the sample surface in the grating measurement process by AFM;

s4, obtaining the actually measured surface topography of the TGZ3 grating sample;

and S5, correcting the actually measured surface morphology of the TGZ3 grating according to the reconstructed AFM pinpoint shape and the line width measurement model to obtain a real grating surface structure.

Further, the AFM tip blind test model is approximately represented by a mathematical morphology development expression:

wherein, I is a point set of the scanned image surface, S is a point set describing sample similarity, and T is a point set describing tip state reflection;

the AFM tip shape estimation function is as follows:

wherein T describes the shape of the needle tip, and d represents the position variation of T in I; the shape of the tip can be expressed by the following formula:

wherein T is_i+1Is the ith result T_i' the (i +1) th iteration result, T_i' (x) is T_iAt x, bringing I into contact with the points contained in I.

Further, step S21, obtaining an AFM image from the profile of the sample scanned by the given tip, including three local maximum points, labeled 1,2 and 3, respectively;

step S22: independently taking two needle points obtained in image areas near the maximum values 1 and 2, then calculating the current local needle point shape through a formula (3), and then taking the intersection of two derived needle points (1 and 2);

step S23: the reconstructed AFM tip shape is obtained by merging the derived tip 3 with the previous results.

Further, the step S3 is specifically:

step S31, to obtain a line width measurement model, firstly defining the shape of the AFM tip as a cone with two unequal semi-cone angles α and β, wherein the side generatrix of the cone is the radius R of the sphere, a is the total cone vertex angle of the outer side generatrix, and delta is the installation angle of the AFM probe;

step S32, introducing the position state parameter of the AFM tip moving along the scanning direction in the sample structure, wherein phi represents the tip scanning inclination angle when the AFM tip scans on the grating;

and step S33, obtaining a geometric relation between the line width profile and the tip through a profile image with a sharp edge structure obtained through AFM signals, wherein the formula is as follows:

a＝α+β (8)

L₍₀₎considering that the probe head affects the measured scanning line width in order to have a height of 0; l is_(h)Considering the probe head influence on the measured scanning line width for the height h;

representing the tip scan tilt angle when the AFM tip is scanned over the grating; w_LineThe real line width is obtained; theta is an included angle between the scanning path and the vertical line width; r is the radius of the probe head; a is the total cone vertex angle of the outside bus;

because the edge radius R of the sharp edge is less than 1nm and far less than the tip radius R (R > R), the edge radius R can be ignored, the direction of the structure line and the scanning direction are not perpendicular when AFM is used for measuring the side wall of the line width structure, the transverse distance of the line width structure in the image does not represent the real line width, and the real line width W is the real line width_LineCan be expressed as equation (9):

W_Line＝L₀cosθ (9)

where θ represents an angle between the scanning direction and the structure line.

And finally, a line width measurement model can be constructed according to comparison between the real geometric shape of the cross section of the tested line width structure and the shape obtained from the AFM and the relationship between the line width and the probe tip particles.

Compared with the prior art, the invention has the following beneficial effects:

the method can effectively reduce and eliminate the distortion of the AFM image by reducing the possible deviation of the AFM image and the shape of an actual sample due to the shape and the position state of the AFM tip under the condition of not changing the hardware facility of the probe, thereby obtaining a more accurate AFM image.

Drawings

FIG. 1 is a schematic flow diagram of the process of the present invention;

FIG. 2 is a schematic diagram of a blind reconstruction method for a needle tip according to an embodiment of the present invention;

FIG. 3 is a schematic diagram illustrating a flow of MATLAB-based blind reconstruction of a needle tip according to an embodiment of the present invention;

FIG. 4 illustrates a conventional scan path in accordance with an embodiment of the present invention;

FIG. 5 is a line width measurement model of a grating sample in accordance with an embodiment of the present invention;

FIG. 6 is a comparison of the true geometry of a cross-section of a test line width structure with a shape obtained from an AFM in an embodiment of the present invention;

fig. 7 is a graph of line width versus probe tip particle for an embodiment of the present invention.

Detailed Description

The invention is further explained below with reference to the drawings and the embodiments.

Referring to fig. 1, the present invention provides a grating atomic force microscope image reconstruction method based on needle point blind reconstruction, including the following steps:

step S1: obtaining the information of AFM (atomic force microscope) needle point morphology by using an OP glass surface or a TGZ3 grating as a needle point characterization device;

step S2, carrying out tip blind reconstruction by using MATLAB according to the obtained AFM tip morphology information to obtain a reconstructed AFM tip shape;

referring to fig. 2, the tip blind test model of AFM is approximated by a mathematical morphology development expression:

wherein, I is a point set of the scanned image surface, S is a point set describing sample similarity, and T is a point set describing tip state reflection;

the AFM tip shape estimation function is as follows:

wherein T describes the shape of the needle tip, and d represents the position variation of T in I; the shape of the tip can be expressed by the following formula:

wherein T is_i+1Is the ith result T_i' the (i +1) th iteration result, T_i' (x) is T_iAt x, bringing I into contact with the points contained in I.

Step S21a, obtaining an AFM image from the profile of a given tip-scanned sample, including three local maximum points, respectively labeled 1,2 and 3;

step S22 a: independently taking two needle points obtained in image areas near the maximum values 1 and 2, then calculating the current local needle point shape through a formula (3), and then taking the intersection of two derived needle points (1 and 2);

step S23 a: the reconstructed AFM tip shape is obtained by merging the derived tip 3 with the previous results.

FIG. 3 is a flow chart of blind reconstruction of acquired AFM probe characterization information based on MATLAB, a detailed operation on the steps of FIG. 2;

the flow chart is shown in FIG. 3, where image I represents tip features, represented by a simplified 9 × 9 matrix;

step S21b, initial probe T₀Arranged as a 3 × 3 matrix, I_iIs a sub-image of 3 × 3 in size in the AFM image;

step S22b selecting sub-graph T by sliding the sampling template (3X 3) over the AFM image_i(i ═ 1,2,. n); if sub-graph I_iIs the largest, the largest pixel is associated with the tip shape T_i-1May be calculated, the difference value is added to the sub-image I_iAfter each predetermined value of (A), I is obtained_i' (i ═ 1, 2.., n), new tip configuration T_iAn element of (i ═ 1, 2.., n) can be determined by combining T with T_i' 1, 2.. n) pixel values are compared with the previous tip shape pixel values to select the smaller one to obtain T_i-1(i ═ 1,2,. n); otherwise, if subgraph I_iIs not the maximum, there is no need to update the previous cusp shape T_i-1(i ═ 1,2,. n), andsliding the sliding template (3 × 3) on the AFM image to the next position;

step S23b of finding all sub-images from the AFM image with the largest center pixel and repeating to update AFM tip T_iAfter the shape of (i ═ 1, 2., n), the final AFM tip shape was obtained.

S3, constructing a line width measurement model according to the geometric relationship between the probe and the sample surface in the grating measurement process by AFM;

FIG. 4 shows a conventional scan path in actual linewidth measurement, under which the probe moves from top to bottom from left to right across the sample surface, with the X direction being the fast scan direction and the Y direction being the slow scan direction, according to a linewidth measurement model, the shape of the AFM tip can be defined as a cone with two unequal half cone angles α and β, with the side generatrices being the radius R of the sphere, a being the total cone apex angle of the outside generatrices, and δ being the mounting angle of the AFM probe. FIG. 5 illustrates an established raster profile measurement model that introduces positional state parameters of the AFM tip as it moves in the scan direction in the sample structure, where φ represents the tip scan tilt angle as the AFM tip scans over the grating. A profile image with a sharp edge structure obtained by AFM signals is shown in FIG. 6, which describes the geometric relationship between the linewidth profile and the tip, and according to the linewidth measurement model, the relationship between these parameters can be expressed as:

a＝α+β (8)

L₍₀₎considering that the probe head affects the measured scanning line width in order to have a height of 0; l is_(h)Considering the probe head influence on the measured scanning line width for the height h;representing the tip scan tilt angle when the AFM tip is scanned over the grating; w_LineThe real line width is obtained; theta is an included angle between the scanning path and the vertical line width; r is the radius of the probe head; a is the total cone vertex angle of the outside bus;

because the edge radius R of the sharp edge is less than 1nm and far less than the tip radius R (R > R), the edge radius R can be ignored, the direction of the structure line and the scanning direction are not perpendicular when AFM is used for measuring the side wall of the line width structure, the transverse distance of the line width structure in the image does not represent the real line width, and the real line width W is the real line width_LineCan be expressed as equation (9):

W_Line＝L₀cosθ (9)

where θ represents an angle between the scanning direction and the structure line.

And finally, a line width measurement model can be constructed according to comparison between the real geometric shape of the cross section of the tested line width structure and the shape obtained from the AFM and the relationship between the line width and the probe tip particles.

S4, obtaining the actually measured surface topography of the TGZ3 grating sample;

and S5, correcting the actually measured surface morphology of the TGZ3 grating according to the reconstructed AFM pinpoint shape and the line width measurement model to obtain a real grating surface structure.

The above description is only a preferred embodiment of the present invention, and all equivalent changes and modifications made in accordance with the claims of the present invention should be covered by the present invention.