Image processing methods to solve imaging distortion in atomic force microscopy under high-speed scanning

By reciprocating scanning and optimizing the rising edge data of the trace-retrace curve, the imaging distortion problem of atomic force microscopy during high-speed scanning in tapping mode was solved, achieving efficient and accurate sample morphology imaging.

CN118570086BActive Publication Date: 2026-11-17NORTHEAST FORESTRY UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202410280587.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-03-12
Publication Date
2026-11-17
Estimated Expiration
2044-03-12

AI Technical Summary

Technical Problem

When atomic force microscopes scan samples at high speed in tapping mode, they suffer from severe imaging distortion, resulting in scanning artifacts and affecting imaging accuracy and efficiency.

Method used

Sample surface morphology data is acquired by reciprocating scanning. By optimizing the trace and retrace scan curves, the rising edge of the trace-retrace curve is used to replace the falling edge data to form a new cross-sectional profile, thereby suppressing imaging distortion caused by system response hysteresis.

Benefits of technology

It effectively suppresses imaging distortion in atomic force microscopy under high-speed scanning, ensuring accurate sample morphology images are obtained at high scanning speeds, thus improving imaging accuracy and efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118570086B_ABST
    Figure CN118570086B_ABST
Patent Text Reader

Abstract

This invention discloses an image processing method to solve imaging distortion in atomic force microscopy (AFM) under high-speed scanning. The method includes the following steps: Step 1, determining the operating mode of the AFM; Step 2, scanning the surface morphology of the sample; Step 3, optimizing the offline image of the scanning result. This invention acquires the morphology data of the sample surface through reciprocating scanning, replacing the falling edge of the trace-retrace curve with the rising edge of the trace-retrace curve. Using this image processing method, scanning artifacts caused by system response lag can be effectively suppressed, ensuring that the AFM can obtain more accurate sample morphology images under high-speed scanning.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of nanoscience and technology, and relates to an atomic force microscope (AFM) imaging processing method, specifically an image processing method to solve the imaging distortion of atomic force microscopes under high-speed scanning. Background Technology

[0002] Atomic force microscopy (AFM) is widely used in surface measurement and precision machining due to its high-resolution imaging capabilities, excellent force measurement capabilities, and positioning accuracy. AFM operates in three modes: contact mode, non-contact mode, and impact mode (semi-contact mode). In contact mode, the probe wears quickly, potentially damaging the sample surface. In non-contact mode, the large distance between the scanning probe and the sample surface results in a low signal reception. Impact mode avoids these drawbacks. Current AFM technology is limited by its mechanical structure and imaging principles, resulting in slow imaging speeds. In impact mode, obtaining a typical sample surface morphology image typically takes tens of minutes. However, by scanning the sample surface at high speed, this time can be reduced to a few minutes. High-speed scanning significantly improves efficiency when processing large numbers of samples. Therefore, current methods primarily focus on increasing the scanning speed of AFM to shorten imaging time. However, while high-speed scanning of samples in atomic force microscopy can shorten imaging time, it also introduces image distortion, known as scanning artifacts. Therefore, the problem of imaging distortion caused by high-speed scanning of sample surfaces in tapping mode using atomic force microscopy is a pressing issue that needs to be addressed. Summary of the Invention

[0003] To address the issue of scanning distortion caused by system response lag when atomic force microscopy (AFM) scans samples at high speed in impact mode, this invention provides an image processing method to resolve imaging distortion during high-speed scanning. This method utilizes the AFM reciprocating scanning method to acquire data on the sample surface morphology, and then optimizes the obtained surface morphology contour information, effectively suppressing scanning artifacts.

[0004] The objective of this invention is achieved through the following technical solution:

[0005] An image processing method for resolving imaging distortion in atomic force microscopy under high-speed scanning includes the following steps:

[0006] Step 1: Determine the operating mode of the atomic force microscope:

[0007] The working mode of the atomic force microscope is adjusted to the tapping mode, and the sample to be scanned is pre-processed accordingly to meet the measurement requirements.

[0008] Step 2: Scan the surface morphology of the sample:

[0009] Step 2: 1. Set the probe's scanning amplitude and the PID system's gain parameters, and set the scanning speed to achieve the desired imaging efficiency;

[0010] Step 22: Using a reciprocating scanning method, while scanning each cross section of the sample surface, simultaneously obtain the contour morphology curve obtained by a trace scan and the contour morphology curve obtained by a retrace scan, and collect the data of the two contour curves at the same time.

[0011] Step 3: Optimize offline images of the scan results:

[0012] Step 3: When processing the first cross section of the sample surface, determine the rising and falling edges of the trace and retrace contour topography scans. Then, extract the rising edge data from the trace and retrace contours respectively, replace the falling edge data of the trace with the rising edge data of the retrace, or replace the falling edge data of the retrace with the rising edge data of the trace, thereby combining them to form a new cross section contour.

[0013] Step 32: Following the method in Step 31, process the other cross sections of the sample surface sequentially to obtain new two-dimensional topographic data of the sample surface.

[0014] Compared with the prior art, the present invention has the following advantages:

[0015] This invention acquires sample surface morphology data through reciprocating scanning, replacing the falling edge of the trace-retrace curve with the rising edge. This image processing method effectively suppresses scanning artifacts caused by system response lag, ensuring that atomic force microscopy can obtain more accurate sample morphology images at high speeds. Attached Figure Description

[0016] Figure 1 Image processing flowchart to address imaging distortion in atomic force microscopy under high-speed scanning;

[0017] Figure 2 This is a schematic diagram illustrating the principle of scanning from the rising edge of the sample morphology using an atomic force microscope.

[0018] Figure 3 This is a schematic diagram illustrating the principle of scanning to the falling edge of the sample morphology using an atomic force microscope.

[0019] Figure 4 This is a comparison of the trace curve and retrace curve of a standard step morphology sample scanned at a scan speed of 40 μm / s.

[0020] Figure 5 This is a comparison of the trace curve and retrace curve of a standard step morphology sample scanned at a scan speed of 120 μm / s.

[0021] Figure 6 This is a schematic diagram showing how to retain line segments in the trace and retrace curves when the scan speed is 120 μm / s.

[0022] Figure 7 A comparison of cross-sectional scanning curves of standard step samples;

[0023] Figure 8 This is a comparison of the trace curve and retrace curve of a sample with a general morphology when the scanning speed is 40 μm / s.

[0024] Figure 9 This is a comparison of the trace curve and retrace curve of a sample with a general morphology when the scan speed is 54.4 μm / s.

[0025] Figure 10 This is a schematic diagram showing how to retain line segments in the trace and retrace curves when the scan speed is 54.4 μm / s.

[0026] Figure 11 The image shows a comparison of the trace curves for scanning speeds of 40 μm / s and 54.4 μm / s.

[0027] Figure 12 The image shows a comparison between the trace curve of a sample with a general morphology at a scan speed of 40 μm / s and the curve of a sample with a general morphology at an optimized scan speed of 54.4 μm / s.

[0028] In the figure: 1 is the atomic force microscope probe; 2 is the amplitude of the probe before scanning the rising edge morphology; 3 is the standard step morphology sample; 4 is the amplitude after the probe amplitude abruptly changes when scanning the rising edge morphology; 5 is the amplitude after the probe amplitude returns to the set value; 6 is the atomic force microscope probe; 7 is the probe amplitude before scanning the falling edge morphology; 8 is the amplitude during the dynamic adjustment process of the probe amplitude when scanning the falling edge of the sample; 9 is the probe amplitude after adjustment; 10 is the standard step morphology sample; 11 is the standard step morphology sample at a scanning speed of 40 μm / s. The retrace scan curves are shown in Figure 12 (for a standard step morphology sample at a scan speed of 40 μm / s), Figure 13 (for a standard step morphology sample at a scan speed of 120 μm / s), Figure 14 (for a standard step morphology sample at a scan speed of 120 μm / s), Figure 15 (for a standard step morphology sample at a scan speed of 120 μm / s), Figure 16 (for a standard step morphology sample at a scan speed of 120 μm / s), and Figure 17 (for a standard step morphology sample at a scan speed of 120 μm / s). The following are trace scan curves of samples at a scan speed of 40 μm / s: 18 is the optimized scan curve of a standard stepped morphology sample at a scan speed of 120 μm / s; 19 is the retrace scan curve of a sample with general morphology at a scan speed of 40 μm / s; 20 is the trace scan curve of a sample with general morphology at a scan speed of 40 μm / s; 21 is the retrace scan curve of a sample with general morphology at a scan speed of 54.4 μm / s; 22 is the trace scan curve of a sample with general morphology at a scan speed of 54.4 μm / s. The following are the trace curves: 23 shows the line segment to be retained in the trace curve at a scan speed of 54.4 μm / s; 24 shows the line segment to be retained in the retrace curve at a scan speed of 54.4 μm / s; 25 shows the trace curve plotted at a scan speed of 40 μm / s; 26 shows the trace curve plotted at a scan speed of 54.4 μm / s; 27 shows the trace scan curve of a sample with general morphology at a scan speed of 40 μm / s; and 28 shows the optimized scan curve of a sample with general morphology at a scan speed of 54.4 μm / s. Detailed Implementation

[0029] The technical solution of the present invention will be further described below with reference to the accompanying drawings, but it is not limited thereto. Any modifications or equivalent substitutions to the technical solution of the present invention that do not depart from the spirit and scope of the technical solution of the present invention should be covered within the protection scope of the present invention.

[0030] Currently, in the use of atomic force microscopes (AFM), to ensure imaging accuracy, the vibrating probe needs to scan the sample surface at a relatively low scanning speed, which results in excessively long scanning times. However, when the probe scans the sample at high speed, the lag in the response of the AFM's PID control system causes the probe to fail to effectively track the undulations of the surface contour, leading to imaging distortion and scanning artifacts in the obtained scan results. Specifically, when the probe scans the rising edge of a step on the sample surface, the tapping amplitude instantly decreases to a constant error signal related to the step height. At this point, the control system responds by restoring the probe's amplitude to the set value, thus obtaining relatively accurate step height information. However, when the probe scans the falling edge of the step, the tapping amplitude does not instantly increase to a constant error signal related to the step height. Instead, it first increases towards the amplitude of free vibration, while the system responds by bringing the probe and sample closer together to maintain the set amplitude until the probe touches the bottom of the step, at which point the dynamic adjustment process ends. During this process, the amplitude error signal is constantly changing, causing a lag in the system response. Therefore, the falling edge contains more error information. The faster the system's scanning speed, the longer the distance the probe travels during the dynamic adjustment process from the top to the bottom of the step, ultimately resulting in a greater "broadening effect" of the falling edge. Therefore, the scanning results of the rising edge provide more accurate morphological information than the falling edge and are less affected by scanning speed. Currently, most commercial atomic force microscopy (AFM) scanning systems use a "trace-retrace" bidirectional reciprocating scanning method for the cross-sectional profile of samples. The overlap of the bidirectional scan profiles is compared to evaluate the probe's tracking of the sample morphology. For the same cross-sectional profile, since the scanning directions of trace and retrace are opposite, the rising edge of the trace scan corresponds to the falling edge of the retrace scan, and vice versa. Therefore, the profile can be composed entirely of the rising edges of the trace and retrace scans, or entirely of the falling edges of the trace and retrace scans. Replacing the falling edge portion in the trace-retrace scan image with the rising edge portion suppresses the imaging distortion caused by system response lag, enabling AFM to achieve higher scanning speeds. Therefore, based on the working principle and characteristics of atomic force microscopy's percussion scanning, this invention proposes an image optimization processing method that utilizes accurate information (rising edge) from trace and retrace scans to form the cross-sectional contour, thereby suppressing broadening artifact errors generated during high-speed scanning. Figure 1 As shown, the specific implementation steps are as follows:

[0031] Step 1: Determine the operating mode of the atomic force microscope.

[0032] The working modes of an atomic force microscope are divided into contact mode, non-contact mode, and impact mode. In this invention, the working mode of the atomic force microscope is adjusted to impact mode, and the sample to be scanned is pre-processed accordingly (such as dividing the sample cross-section) to meet the measurement requirements.

[0033] Step 2: Scan the surface morphology of the sample.

[0034] This invention employs a reciprocating scanning method to scan the sample surface. The scanning amplitude of the probe and the gain parameters of the PID system are set, along with the scanning speed, to achieve the desired imaging efficiency. Using the reciprocating scanning method, when scanning each cross-section of the sample surface, both a trace scan and a retrace scan contour curve can be obtained simultaneously, allowing for the simultaneous acquisition of data from both contour curves.

[0035] Step 3: Optimize the offline images of the scan results.

[0036] After scanning the sample surface morphology, the acquired trace and retrace morphology data are processed. When processing the first cross-section of the sample surface, the rising and falling edges of the trace and retrace contours are determined. Then, the rising edge data from the trace and retrace contours are extracted, and the rising edge data of the retrace replaces the falling edge data of the trace, or vice versa, thus combining them to form a new cross-sectional contour. The second, third, and subsequent cross-sections of the sample surface are processed in the same way, resulting in new two-dimensional sample surface morphology data. This data reduces the artifact error caused by the falling edge, ensuring the accuracy of probe measurements under high-speed scanning.

[0037] Example:

[0038] First, a standard step morphology sample is presented, followed by a sample with a general morphology. The atomic force microscope is operated in tapping mode, such as... Figure 2 As shown, when the atomic force microscope (AFM) scans to the rising edge of the step on the surface of a standard step-shaped sample, the amplitude of the AFM probe instantly decreases to a constant error signal related to the step height. At this point, the AFM PID control system immediately adjusts the probe amplitude to restore it to the set value, and the amplitude returns to normal. Figure 3As shown, when the atomic force microscope (AFM) scans to the descending edge of the step on the surface of a standard step-shaped sample, the amplitude of the AFM probe does not instantaneously increase to a constant error signal related to the step height. At this point, the AFM PID control system adjusts the distance between the probe and the sample until the probe contacts the bottom of the standard step-shaped sample, at which point the amplitude returns to normal, and the dynamic adjustment process ends. During this process, the error signal of the probe amplitude changes continuously, which leads to a lag in the system response. Moreover, the faster the scanning speed, the more severe the lag in the system response. The image processing method proposed in this invention suppresses scanning artifacts, enabling the AFM to obtain a true image of the sample surface morphology even at high-speed scanning.

[0039] Implementation Method 1: Scanning the standard step sample

[0040] Step 1: Selecting the working mode of the atomic force microscope and scanning the sample:

[0041] In this embodiment, the atomic force microscope (AFM) used is the Veeco Dimension Icon, with a maximum scanning range of 89 μm × 89 μm × 9.8 μm, a closed-loop position noise of 0.1 nm in the XY direction, and a closed-loop sensor noise level of 35 pm in the Z direction. The probe used is a Bruker RTESP single-crystal silicon probe with an elastic modulus of approximately 40 N / m and a resonant frequency of approximately 330-350 kHz. The scanned sample is a Bruker VGRP-15M calibration sample, a 10 μm × 10 μm square grid array with a height of 180 nm and a platinum coating. The AFM was set to tapping mode, with the integral gain set to 0.5, the proportional gain to 5, and the derivative gain to 1 in the PID control system. The free amplitude was set to 474 mV, and the scan amplitude to 260 mV. The scan length for each cross-section of the sample surface was set to 40 μm, and the number of scan points was 128. The sample was scanned at speeds of 40 μm / s and 120 μm / s, respectively.

[0042] Step 2: Quality evaluation of scanned images:

[0043] The sample scanned in step one is a standard AFM sample manufactured by Bruker, and the sample surface has a 10μm × 10μm square grid array with a height of 180nm. Figure 4As shown, at a scan speed of 40 μm / s, the height and width data of the square grid array measured by trace scan (curve 12) and retrace scan (curve 11) were averaged to calculate the height values ​​as 179.21 nm and 180.46 nm, respectively. Compared with the standard sample, the error rates were 0.4% and 0.26%, respectively. The calculated width was 10.21 μm for both measurements, with an error rate of 2.04%. Figure 5 As shown, at a scanning speed of 120 μm / s, the height and width data of the square grid array measured by trace scan (curve 14) and retrace scan (curve 13) were used to calculate the height values ​​as 182.12 nm and 182.48 nm, respectively, by averaging the values. Compared with the standard sample, the error rates were 1.16% and 1.36%, respectively. The calculated width values ​​were 10.26 μm and 10.32 μm, respectively, with error rates of 2.53% and 3.1%, respectively.

[0044] Step 3: Optimize the image of the sample.

[0045] Based on the data collected in step one, the trace curve (curve 14) and retrace curve (curve 13) when the scan speed is 120μm / s have been plotted in step two. In this step, the trace curve and retrace curve in step two are optimized.

[0046] The first step is to process the trace curve. First, mark the starting and ending points of the curve's descent in the trace curve. Figure 5 The rising and falling trends of the trace curve have been marked. Then, the line segments between the two ending points of the fall, excluding the falling portion of the curve, are retained sequentially. The retained segments consist of: the line segment starting from the ending point closer to the origin and ending at the starting point of the fall between the two ending points (e.g., ...). Figure 6 As shown in curve 15, the retained portion of the trace curve is marked with a circle.

[0047] The second step is to process the retrace curve. First, mark the starting and ending points of the descent in the retrace curve. Figure 5 The rising and falling trends of the retrace curve have been marked. Then, the line segments between the two ending points of the descent, excluding the falling portion of the curve, are retained sequentially. The retained segments consist of: the line segment starting from the ending point further away from the origin and ending at the starting point of the descent between the two ending points (e.g., ...). Figure 6 As shown in curve 16, the retained portion of the retrace curve is marked with a triangle.

[0048] Finally, based on the retained trace and retrace curve segments, coordinate points are extracted, and these points are reordered in ascending order of their x-coordinates. If points with the same x-coordinate exist, their y-coordinates are compared; points with smaller y-coordinates are retained, and points with larger y-coordinates are deleted. The optimized curve image is then redrawn (e.g., ...). Figure 7 (As shown in curve 18).

[0049] Step 4: Verification of optimization results:

[0050] like Figure 7 As shown, curve 17 (trace curve at a scanning speed of 40 μm / s) and curve 18 (optimized curve) have a high degree of overlap. The height of the square grid array calculated from curve 18 is 180.12 nm and the width is 10.11 μm. Compared with the given standard sample, the error rates are 0.07% and 1.07%, respectively, indicating that the data obtained from the optimized curve is closer to the size parameters given by the standard sample.

[0051] Implementation Method 2: Scanning general morphology samples with highly random variations

[0052] Step 1: Selecting the working mode of the atomic force microscope and scanning the sample:

[0053] The sample used was a titanium sample, model RS-15M, manufactured by Bruker. The scanning length of each cross-section of the sample surface was set to 20 μm, and the number of scanning points was 256. The scanning speed was 40 μm / s and 54.4 μm / s, and the sample was scanned twice.

[0054] Step 2: Quality evaluation of scanned images:

[0055] Based on the data acquired in step one, trace curves and retrace curves were plotted at scan speeds of 40 μm / s and 54.4 μm / s, respectively. The overlap between the trace curve and the retrace curve was used to measure the imaging effect of the atomic force microscope, introducing the imaging quality Q:

[0056] Q = ∑|x(trace) - x(retrace)|

[0057] Where x(trace) and x(retrace) are the sample height morphology data obtained when completing one bidirectional reciprocating scan. The smaller Q is, the higher the overlap between the two, reflecting the good tracking effect of the probe on the sample surface, that is, the better the imaging effect of the atomic force microscope, and vice versa.

[0058] like Figure 8 As shown, when the scanning speed is 40 μm / s, the image quality Q can be calculated to be 7.36 × 10⁻⁶ based on curves 19 (retrace scan curve) and 20 (trace scan curve). 3 .like Figure 9 As shown, when the scanning speed is 120 μm / s, the image quality Q can be calculated to be 9.34 × 10⁻⁶ based on curves 21 (retrace scan curve) and 22 (trace scan curve). 3 When the scanning speed is 40 μm / s, the image quality Q is relatively small, resulting in good imaging performance of the atomic force microscope. This indicates that when the scanning speed is 40 μm / s, both the trace curve and the retrace curve can reflect the true surface morphology of the sample.

[0059] Step 3: Optimize the image of the sample:

[0060] Based on the data collected in step one and the trace and retrace curves plotted in step two, the curve image at a scan speed of 54.4 μm / s is optimized. Furthermore, the image optimization method is the same as in implementation method one (step three), and the velocity directions of the trace and retrace curves are in... Figure 9 The above indicates the rising and falling trends of the trace and retrace curves based on the speed direction (see also [reference]). Figure 5 The rising and falling trends of the trace curve and retrace curve, and the retained portion of the trace curve, in Figure 10 The retained portion of the retrace curve is marked with a circle in curve 23. Figure 10 Curve 24 is marked with a triangle.

[0061] Step 4: Verification of optimization results:

[0062] like Figure 11 As shown, curve 25 is the trace curve under slow scanning (scanning speed of 40 μm / s), and curve 26 is the trace curve under fast scanning (scanning speed of 54.4 μm / s). Calculations show that the maximum difference between the two is 88.13, the average difference is 23.33, and the variance is 14.65, indicating that scanning artifacts exist under fast scanning. Figure 12In the figure, curve 27 is the trace curve under slow scanning, and curve 28 is the fast scanning image curve after optimization. The calculation shows that the maximum difference between the two is 33.52, the average difference is 7.90, and the variance of the difference is 5.74. This indicates that by using image optimization processing, scanning artifacts can be suppressed, thereby reflecting the true surface morphology of the sample.

Claims

1. An image processing method for solving the imaging distortion of an atomic force microscope under high-speed scanning, characterized by The method includes the following steps: Step 1: Determine the operating mode of the atomic force microscope: The working mode of the atomic force microscope is adjusted to the tapping mode, and the sample to be scanned is pre-processed accordingly to meet the measurement requirements. Step 2: Scan the surface morphology of the sample: Step 2:

1. Set the probe's scanning amplitude and the PID system's gain parameters, and set the scanning speed to achieve the desired imaging efficiency; Step 22: Using a reciprocating scanning method, while scanning each cross section of the sample surface, simultaneously obtain the contour morphology curve obtained by a trace scan and the contour morphology curve obtained by a retrace scan, and collect the data of the two contour curves at the same time. Step 3: Optimize offline images of the scan results: Step 3: When processing the first cross-section of the sample surface, determine the rising and falling edges of the trace and retrace contours during scanning. Then, extract the rising edge data from the trace and retrace contours respectively. Replace the falling edge data of the trace with the rising edge data of the retrace, or replace the falling edge data of the retrace with the rising edge data of the trace, thereby combining them to form a new cross-sectional contour. The specific steps are as follows: Step 3: Processing the trace curve: First, mark the starting point and ending point of the descent of the curve in the trace curve. Then, retain the line segments between the two ending points of descent, excluding the descent part of the curve. That is, the retained line segments are composed of the following: between the two ending points of descent, starting from the ending point of descent closer to the origin and ending at the starting point of descent between the two ending points of descent. Step 3.12: Processing the retrace curve: First, mark the starting point and ending point of the descent of the retrace curve. Then, sequentially retain the line segments between the two ending points of descent, excluding the descent portion of the curve. That is, the retained line segments are composed of the following: between the two ending points of descent, starting from the ending point of descent further away from the origin and ending at the starting point of descent between the two ending points of descent. Step 313: Based on the retained trace curve segment and retrace curve segment, extract the coordinate points and reorder the extracted coordinate points in order of increasing x-coordinate. If there are coordinate points with the same x-coordinate, compare the y-coordinate values ​​of the same coordinate points, keep the coordinate points with smaller y-coordinate values, delete the coordinate points with larger y-coordinate values, and redraw the optimized curve image. Step 32: Following the method in Step 31, process the other cross sections of the sample surface sequentially to obtain new two-dimensional topographic data of the sample surface.

Citation Information

Patent Citations

  • Atomic force microscope dynamic imaging method based on step response curve of piezoelectric scanning tube

    CN103472266A

  • Intelligent high-speed scanning method and device for spot scanning long-life fluorescence microscopic imaging

    CN113504634A