Hydrophobic micro / nano-scale flexible thin film structure and characterization method

Abstract

This invention discloses a hydrophobic micro / nano-scale flexible thin film structure and its characterization method, belonging to the field of thin film structure characterization technology. To address the problems of poor support and inaccurate standard data acquisition in flexible thin films, this invention sets different accelerating voltages and currents for the coarse and fine thinning stages, enabling precise material removal and ensuring accurate thinning thickness. Using a ruler tool, the thickness, surface undulation amplitude, and microstructural feature dimensions of the thin film in the measurement image are measured, converting the morphological characteristics of the hydrophobic micro / nano-scale flexible thin film into precise data. The high hardness, high chemical stability, and good thermal conductivity of tungsten provide rigid support for the thin film. First, the upper and lower sides of the target area are milled and the surrounding material is removed to form a preliminary thin film outline. Then, U-shaped milling and backsplashing material cleaning are performed, effectively controlling the sample morphology and reducing the damage to the sample structure caused by processing stress.

Description

Technical Field

[0001] This invention relates to the field of thin film structure characterization technology, specifically to hydrophobic micro / nano-scale flexible thin film structures and characterization methods. Background Technology

[0002] Flexible films have important applications in consumer electronics, automotive electronics, medical devices, smart textiles, aerospace, flexible solar cells, flexible sensors, and flexible batteries.

[0003] Chinese patent CN116432431A discloses a fractal growth modeling and simulation method for the surface morphology of flexible silver nanofilms. It mainly simulates the fractal growth of the surface morphology of flexible silver nanofilms under different preparation process parameters by changing model parameters. The simulation results guide experimental research on the preparation of flexible metal nanofilms, accelerating the optimization of the preparation process and improving the efficiency of laboratory research. It can run computer simulation calculations to quickly simulate the surface morphology of flexible silver nanofilms prepared by in-situ reduction method. Although the above patent solves the problem of obtaining thin film characterization data, the following problems still exist in practical operation:

[0004] 1. Direct gold evaporation treatment on the surface of flexible films results in gold particles embedding into the surface of the flexible films, which damages the surface morphology of the flexible films.

[0005] 2. The flexible film was not specifically processed and protected, which made the film surface prone to deformation.

[0006] 3. The lack of targeted thickness removal for the flexible film and the absence of effective measurement of the final flexible film resulted in inaccurate final film data. Summary of the Invention

[0007] The purpose of this invention is to provide a hydrophobic micro / nano-scale flexible thin film structure and a characterization method. By setting different accelerating voltages and currents for the coarse and fine thinning stages, precise material removal can be achieved, ensuring the accuracy of the thinning thickness. Using a ruler tool, the thickness, surface undulation amplitude, and microstructure feature size of the thin film in the measurement image can be measured, converting the morphological characteristics of the hydrophobic micro / nano-scale flexible thin film into accurate data. The high hardness, high chemical stability, and good thermal conductivity of tungsten provide rigid support for the thin film. First, the upper and lower sides of the target area are milled and the surrounding material is punched to form a preliminary thin film outline. Then, U-shaped milling and anti-splashing material cleaning are performed to effectively control the sample morphology and reduce the damage of processing stress to the sample structure, thus solving the problems in the prior art.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] Hydrophobic micro / nano-scale flexible thin film structures include flexible thin film body structure, surface protection structure, conductive structure, support structure and fixing structure;

[0010] The surface protection structure is formed by lightly drawing on the thin film surface with an oil-based marker, creating dotted ink distribution due to hydrophobicity, which is used to mark the sampling area and protect the surface; the conductive structure is a gold vapor deposition layer; the support structure includes an upper protective layer, a sample layer, and a lower protective layer; and the fixing structure is a copper grid substrate.

[0011] Characterization methods for hydrophobic micro / nano-scale flexible thin film structures include:

[0012] First, the thin film sample is pre-processed. The pre-processed sample is then marked with the target area and initially protected. The initially protected sample is then rough-processed. After rough processing, the sample is separated and then protected a second time. The second-protected sample is then thinned and prepared. After thinning, microstructure imaging is performed under low-damage conditions to obtain thin film size and morphology data.

[0013] Preferably, the thin film sample undergoes sample pretreatment, including:

[0014] Draw a line on the surface of the hydrophobic film with an oil-based marker. After drawing the line, the film surface will be dotted and form a protective surface structure.

[0015] Let the surface of the film stand until the ink dries;

[0016] The dried film sample is placed in the gold vapor deposition equipment, and a gold film is vapor deposited on the surface of the film sample.

[0017] Finally, the sample pretreatment of the thin film sample was completed.

[0018] Preferably, the pretreated sample is marked with target regions, and the marked target regions are initially protected, including:

[0019] The pretreated thin film sample was placed in the sample chamber of a dual-beam electron microscope with a focused ion beam and an electron beam, and the sample surface was scanned at low magnification using an electron beam.

[0020] Based on the scanning results, the dotted traces on the sample surface are identified;

[0021] The identified point-like distribution area is taken as the target area, and the boundary of the target area is marked using the marking function of the focused ion beam electron beam dual-beam electron microscope.

[0022] Next, the target area is initially protected by adjusting the ion beam angle of the focused ion beam electron beam dual-beam electron microscope, wherein the ion beam angle is 52°, the accelerating voltage is set to 30kV, and the ion beam current is set to 0.23nA.

[0023] A tungsten protective layer with a thickness of 1 μm was deposited on the surface of the target area;

[0024] The deposition process is initiated, and tungsten material covers the target area under the action of an ion beam, forming the initial structure of the upper protective layer and the sample layer.

[0025] Preferably, the identification of dot-like traces on the sample surface based on the scanning results includes:

[0026] The electron micrographs obtained by scanning the sample surface are preprocessed to obtain preprocessed electron micrographs; wherein, the preprocessing includes grayscale normalization and noise suppression.

[0027] Retrieve the global grayscale mean and grayscale standard deviation of the preprocessed electron micrograph;

[0028] A dynamic threshold is set using the global grayscale average value and grayscale standard deviation of the preprocessed electron micrograph;

[0029] The pixels contained in the submicroscopic image are compared with a dynamic threshold, and pixels with gray values ​​exceeding the dynamic threshold are selected as candidate pixels.

[0030] Resonance scoring is performed on candidate pixels to obtain the resonance score value corresponding to each candidate pixel;

[0031] The resonance score value is compared with a preset score threshold, wherein the preset score threshold ranges from 0.62 to 0.67.

[0032] Candidate pixels whose resonance score values ​​exceed a preset score threshold are selected as target pixels.

[0033] For each target pixel, a trace probability evaluation index is generated using the resonance score and the pixel confidence level.

[0034] The trace probability evaluation index corresponding to the target pixel is compared with a preset index threshold; wherein the preset index threshold ranges from 0.42 to 0.51.

[0035] Target pixels whose trace probability evaluation index is not lower than a preset index threshold are taken as valid trace pixels, and the valid trace pixels are integrated to form a dotted distribution area.

[0036] Preferably, resonance scoring processing is performed on candidate pixels to obtain the resonance score value corresponding to each candidate pixel, including:

[0037] The preprocessed electron microscopy image is subjected to multi-scale feature extraction to obtain multi-scale features; wherein, the multi-scale features include spatial scale features, material intrinsic features, geometric topological features and dynamic process features;

[0038] Local features of each candidate pixel are extracted, wherein the local features of the candidate pixel include gray-level distribution, texture pattern and physical field parameters;

[0039] Determine the cosine similarity S(x, y) between the feature vector formed by the local features of each candidate pixel and the feature vector formed by the multi-scale features.

[0040] Calculating the phase gradient Φ from electron micrographs BSE ;

[0041] A topology network is established on the candidate pixels, and a topology continuity score is calculated for each candidate pixel in the topology network.

[0042] The resonance score corresponding to candidate pixel (x, y) is obtained by combining the topological continuity score of each candidate pixel with the geometric constraint factor of each candidate pixel.

[0043] Preferably, the sample after preliminary protection is subjected to rough processing, followed by sample separation, including:

[0044] Before roughing, the parameters of the focused ion beam electron beam dual-beam electron microscope are set, including adjusting the ion beam angle of the focused ion beam electron beam dual-beam electron microscope to 52°, the accelerating voltage to 30kV, and the ion beam current to 9.3nA.

[0045] Centered on the sample that has been initially protected, the upper and lower sides of the target area are milled using an ion beam, and the surrounding material of the target area is hollowed out, forming a preliminary thin sheet outline after hollowing out.

[0046] After forming the initial thin-film outline, the ion beam angle is changed to 0°, the voltage is maintained at 30kV, the current is reduced to 2.5nA, and U-shaped milling is performed on the bottom and sides of the target area.

[0047] After U-shaped milling, the ion beam angle is restored to 52°, the voltage is 30kV, and the current is 2.5nA to clean up the splashed material generated during the U-shaped milling process;

[0048] After cleaning, the sample undergoes rough processing, followed by sample separation.

[0049] Sample separation involves first switching the focused ion beam electron beam dual-beam microscope to low current mode, with an ion beam angle of 0° and a voltage of 30kV. The micromanipulation needle is then lowered above the target area of ​​the rough-processed sample, so that the tip of the micromanipulation needle contacts the surface of the target area. Carbon is deposited by ion beam induction to adhere and fix the micromanipulation needle to the target area.

[0050] After adhesion, lift the micromanipulation needle upwards to detach it from the substrate, then rotate it 180° so that the tip of the micromanipulation needle is pointing downwards and the bottom of the target area is facing upwards. Place the rotated sample needle tip into the sample stage, connect the sample to the sample stage using carbon, and then sever the connection between the micromanipulation needle and the sample.

[0051] Rotate the sample stage 180° again so that the bottom of the target area is facing upwards. Lower the micromanipulation needle again to contact the target area. After connecting with carbon, disconnect the sample from the sample stage and extract the sample back onto the micromanipulation needle. Rotate the needle 180° again to restore the sample to its normal orientation.

[0052] Finally, move the micromanipulation needle above the copper grid, adjust its position so that the target area is aligned with the copper grid mesh, use carbon to bond the sample to the side of the copper grid, then use an ion beam to cut the connection between the sample and the micromanipulation needle, pull out the needle, and complete the fixation of the sample on the copper grid.

[0053] The separation of the samples was finally completed.

[0054] Preferably, the separated samples undergo secondary protection, including:

[0055] The ion beam angle of the focused ion beam electron beam dual-beam electron microscope was adjusted to 0°, the accelerating voltage was set to 30kV, and the ion beam current was reduced to 80pA.

[0056] Initiate the ion beam-induced deposition process to deposit a 1 μm thick protective layer on the lower surface of the sample fixed on the copper grid using tungsten as the material;

[0057] During the deposition process, the ion beam continuously scans the bottom of the sample, ensuring uniform coverage of the tungsten material and forming a continuous tungsten layer;

[0058] After deposition, the sample cross-section was observed in electron beam mode to confirm that the upper protective layer, sample layer and lower protective layer formed a complete three-layer structure. The upper protective layer is the tungsten layer deposited during the initial protection and the lower protective layer is the tungsten layer deposited during the secondary protection.

[0059] Finally, the secondary protection of the sample was completed.

[0060] Preferably, the sample after secondary protection is thinned for sample preparation, including:

[0061] The thinned area of ​​the sample after secondary protection was confirmed;

[0062] After confirming the thinning region, the parameters of the focused ion beam and electron beam dual-beam electron microscope were adjusted. Specifically, the ion beam angle was set to 52°, and the accelerating voltage was divided into a coarse thinning stage and a fine thinning stage. The current in the coarse thinning stage was set to 0.43 nA, and the current in the fine thinning stage was reduced to 80 pA.

[0063] The thinning area is first subjected to coarse thinning, which is as follows: the thinning area is milled with an ion beam angle of 52° and a current of 0.43nA, and the material is removed layer by layer from the top protective layer to the bottom until the remaining thickness is 100nm. When the milling reaches the preset time, the SEM mode is switched to scan and the remaining thickness is measured.

[0064] The thinned area is then further thinned by reducing the ion beam current to 80 pA and milling the thinned area. The thickness of the sample is measured after each milling until the sample thickness reaches 50 nm.

[0065] After the sample is thinned, the sample surface is scanned with an ion beam. If local thickness unevenness is found after scanning, an 80pA current is used to refine the thick area.

[0066] Then clean the surface of the sample that has been fixed to the copper grid and thinned.

[0067] Finally, the sample thinning and preparation were completed.

[0068] Preferably, after thinning and sample preparation, microstructure imaging is performed under low-damage conditions to obtain film size and morphology data, including:

[0069] The thinned sample is loaded into the sample holder of the focused ion beam electron beam dual-beam electron microscope.

[0070] In a dual-beam focused ion beam electron beam microscope, the edge of the copper grid or a non-target area is first selected as the focal point. The sample is coarsely focused by low magnification. At the same time, the electron beam avoids the target area of ​​the sample during the focusing process.

[0071] Then, the target area of ​​the sample on the copper grid is located by low-magnification scanning. At the same time, the magnification is increased to 10000× to confirm the sample position.

[0072] Fine focusing is performed in a blank area far from the target area of ​​the sample. After focusing, the target area of ​​the sample is moved quickly by moving the sample stage to the center of the field of view.

[0073] After the sample is translated, a high-sensitivity CCD camera is used to expose the sample. If multiple areas are observed, the field of view is switched quickly in sequence, and each area is photographed no more than twice to obtain the final measurement image.

[0074] The thickness, surface undulation amplitude, and microstructure feature size of the thin film in the measurement image are measured using a ruler tool;

[0075] The final measurement data are characterization data for hydrophobic micro / nano-scale flexible films.

[0076] Compared with the prior art, the present invention has the following beneficial effects:

[0077] 1. The hydrophobic micro / nano-scale flexible thin film structure and characterization method provided by the present invention can significantly improve the conductivity of the thin film sample by depositing a gold film on the surface of the thin film sample, avoid the impact of charge accumulation on the imaging quality during detection by electron microscopy and other processes, and the gold film provides an additional protective barrier for the thin film sample, further preventing the sample from being oxidized or contaminated and extending the sample preservation time.

[0078] 2. The hydrophobic micro-nano-scale flexible thin film structure and characterization method provided by the present invention first mills the upper and lower sides of the target area and drills the surrounding material to form a preliminary thin film outline, and then performs U-shaped milling and anti-splash material cleaning. This step-by-step processing method can effectively control the sample morphology and reduce the damage of processing stress to the sample structure. The high hardness, high chemical stability and good thermal conductivity of tungsten provide rigid support for the thin film.

[0079] 3. The hydrophobic micro / nano-scale flexible thin film structure and characterization method provided by this invention, with its operation process of coarse thinning followed by fine thinning, makes the sample preparation process more orderly and controllable. By setting different accelerating voltages and currents for the coarse and fine thinning stages, the material can be precisely removed, ensuring the accuracy of the thinning thickness. By using a ruler tool to measure the thickness, surface undulation amplitude, and microstructure feature size of the thin film in the measurement image, the morphological characteristics of the hydrophobic micro / nano-scale flexible thin film can be converted into accurate data. Attached Figure Description

[0080] Figure 1 This is a schematic diagram of the characterization steps for the hydrophobic micro / nano-scale flexible thin film of the present invention. Detailed Implementation

[0081] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0082] To address the problem in existing technologies where gold vapor deposition directly affects the surface of flexible films, leading to gold particles embedding into the film and disrupting its surface morphology, please refer to [link to relevant documentation]. Figure 1This embodiment provides the following technical solution:

[0083] Hydrophobic micro / nano-scale flexible thin film structures include flexible thin film body structure, surface protection structure, conductive structure, support structure and fixing structure;

[0084] The surface protection structure is formed by lightly drawing on the thin film surface with an oil-based marker, creating dotted ink distribution due to hydrophobicity, which is used to mark the sampling area and protect the surface; the conductive structure is a gold vapor deposition layer; the support structure includes an upper protective layer, a sample layer, and a lower protective layer; and the fixing structure is a copper grid substrate.

[0085] Specifically, the surface protection structure utilizes hydrophobicity to allow oil-based markers to form dotted ink distribution, which can accurately mark the sampling area, avoiding scratches and contamination of the film surface caused by traditional marking methods. Furthermore, its hydrophobic properties isolate external moisture, dust, and other impurities, preventing the film surface from altering its properties due to foreign matter adsorption. This effectively improves the film's durability and surface cleanliness in complex environments, providing reliable initial conditions for subsequent sample analysis and experimental operations. The gold vapor deposition layer, as a conductive structure, leverages the excellent conductivity, chemical stability, and oxidation resistance of gold to ensure efficient and stable current transmission at the micro-nano scale, reducing resistance loss and signal attenuation. The support structure consists of an upper protective layer, a sample layer, and a lower protective layer. These three layers work synergistically to provide excellent mechanical support and buffer protection for the film. The copper grid substrate, as a fixing structure, possesses good mechanical strength and processing performance, facilitating connection and fixation with other equipment or carriers.

[0086] Characterization methods for hydrophobic micro / nano-scale flexible thin film structures include:

[0087] First, the thin film sample is pre-processed. The pre-processed sample is then marked with the target area and initially protected. The initially protected sample is then rough-processed. After rough processing, the sample is separated and then protected a second time. The second-protected sample is then thinned and prepared. After thinning, microstructure imaging is performed under low-damage conditions to obtain thin film size and morphology data.

[0088] Specifically, marking and initial protection of the target area can precisely locate hydrophobic functional regions or special micro / nano structure sites on the film surface, avoiding information redundancy caused by global processing. The dual protection mechanism (initial protection and secondary protection) and low-damage sample preparation process can effectively address the easily deformable and damaged characteristics of flexible films. In the rough processing and separation stages, initial protection can prevent mechanical forces from pulling and damaging the film microstructure; secondary protection, combined with low-damage treatment during thinning sample preparation (such as ion beam thinning or vibration polishing), can avoid the damage to hydrophobic interfaces or flexible substrates caused by high temperature and high pressure in traditional sample preparation. The whole process control from pretreatment to low-damage imaging can minimize artifacts introduced by sample preparation. Pretreatment removes surface contaminants (such as residual solvents and dust) to avoid interference from impurities in morphology observation; thinning the sample ensures that the film meets the observation thickness requirements of high-precision equipment such as transmission electron microscopes (usually <100nm), ensuring that the three-dimensional information of the micro-nano structure can be accurately restored through two-dimensional imaging; low-damage imaging (such as using low-voltage scanning electron microscopes or cryo-electron microscopes) can avoid the collapse of the film structure or changes in chemical composition caused by electron beam irradiation, making the obtained dimensions (such as nanopillar height and pore diameter) and morphology data (such as surface roughness and structural periodicity) closer to the true state of the sample.

[0089] Sample pretreatment of the thin film sample includes:

[0090] Draw a line on the surface of the hydrophobic film with an oil-based marker. After drawing the line, the film surface will be dotted and form a protective surface structure.

[0091] Let the surface of the film stand until the ink dries;

[0092] The dried film sample is placed in the gold vapor deposition equipment, and a gold film is vapor deposited on the surface of the film sample.

[0093] Finally, the sample pretreatment of the thin film sample was completed.

[0094] Specifically, using an oil-based marker to draw a line on the surface of the hydrophobic thin film creates a dotted distribution and surface protection structure. This operation is simple and efficient. The dotted distribution increases the surface roughness of the film, altering its surface properties and making it easier for the film to interact with other substances in subsequent experiments, thus improving the sensitivity and stability of the detection signal. Simultaneously, the formed surface protection structure can, to some extent, prevent physical damage to the film surface during subsequent operations, protecting the original characteristics of the film sample and ensuring the reliability of experimental results. Allowing the film surface to dry completely ensures that the oily substance will not interfere with the subsequent gold film deposition step, preventing uneven gold film deposition or failure to adhere due to oil residue. Only after the marker ink is completely dry are the physicochemical properties of the film surface in a stable state, providing a good substrate condition for gold film deposition. Depositing a layer of gold on the surface of the thin film sample has many advantages. Gold has good conductivity and chemical stability, which can significantly improve the conductivity of the thin film sample and avoid the impact of charge accumulation on imaging quality during electron microscopy and other detection processes. Meanwhile, the gold film provides an additional protective barrier for the thin film sample, further preventing oxidation or contamination and extending its shelf life. Furthermore, the presence of the gold film enhances the reflectivity of the thin film sample surface, improving the accuracy and clarity of detection in experiments such as optical inspection.

[0095] The pretreated sample is then marked with the target region, and the marked target region is initially protected, including:

[0096] The pretreated thin film sample was placed in the sample chamber of a dual-beam electron microscope with a focused ion beam and an electron beam, and the sample surface was scanned at low magnification using an electron beam.

[0097] Based on the scanning results, the dotted traces on the sample surface are identified;

[0098] The identified point-like distribution area is taken as the target area, and the boundary of the target area is marked using the marking function of the focused ion beam electron beam dual-beam electron microscope.

[0099] Next, the target area is initially protected by adjusting the ion beam angle of the focused ion beam electron beam dual-beam electron microscope, wherein the ion beam angle is 52°, the accelerating voltage is set to 30kV, and the ion beam current is set to 0.23nA.

[0100] A tungsten protective layer with a thickness of 1 μm was deposited on the surface of the target area;

[0101] The deposition process is initiated, and tungsten material covers the target area under the action of an ion beam, forming the initial structure of the upper protective layer and the sample layer.

[0102] Specifically, by using a focused ion beam-electron beam dual-beam electron microscope (FBM) to scan the sample surface at low magnification, a comprehensive image of the sample can be quickly acquired. Combined with the precise identification of point-like traces on the sample surface, target areas can be efficiently located. This method avoids blind searching, significantly improving the accuracy and efficiency of target area marking, laying a solid foundation for subsequent analysis. By adjusting the ion beam angle to 52° and setting the accelerating voltage to 30kV and the ion beam current to 0.23nA, targeted processing of the target area can effectively reduce ion beam damage to the sample. Based on this, a 1μm thick tungsten protective layer is deposited on the target area surface, forming the preliminary structure of the upper protective layer and the sample layer. This isolates the sample from external interference, enhances the structural stability of the sample, and ensures the integrity of the target area during subsequent processing. The dual-beam FBM's dual functions—electron beam scanning and identification, and ion beam marking and protection—fully leverage the advantages of the equipment and achieve efficient collaboration between the technologies. Meanwhile, precisely setting the ion beam parameters and deposition thickness is a deep optimization of the process, ensuring the optimal protection effect, providing a reliable guarantee for subsequent research and analysis of the samples, and improving the overall quality and reliability of the research.

[0103] Preferably, the identification of dot-like traces on the sample surface based on the scanning results includes:

[0104] The electron micrograph (BSE) obtained from scanning the sample surface is preprocessed to obtain a preprocessed electron micrograph; wherein, the preprocessing includes grayscale normalization and noise suppression.

[0105] Retrieve the global grayscale mean and grayscale standard deviation of the preprocessed electron micrograph;

[0106] A dynamic threshold is set using the global grayscale average value and grayscale standard deviation of the preprocessed electron micrograph;

[0107] The dynamic threshold is obtained using the following formula:

[0108]

[0109] Where Y represents the dynamic threshold; μ G σ represents the global average grayscale value. G Represents the grayscale standard deviation; A represents the connected area of ​​a 5×5 pixel neighborhood; specifically, As a connected area-driven correction intensity regulator, it achieves "the more continuous the local structure, the more aggressive the threshold adjustment; the more fragmented the local structure, the more conservative the threshold." For example, when analyzing second-phase particles in an alloy, if A is large inside the particle, the correction term makes the threshold adapt to the grayscale inside the particle; if A is small at the particle edge, the threshold is more sensitive, accurately separating the particle from the matrix.

[0110] Each pixel in the electron micrograph is compared with a dynamic threshold, and pixels with gray values ​​exceeding the dynamic threshold are selected as candidate pixels.

[0111] Resonance scoring is performed on candidate pixels to obtain the resonance score value corresponding to each candidate pixel;

[0112] The resonance score value is compared with a preset score threshold, wherein the preset score threshold ranges from 0.62 to 0.67.

[0113] Candidate pixels whose resonance score values ​​exceed a preset score threshold are selected as target pixels.

[0114] For each target pixel, a trace probability evaluation index is generated using the resonance score and the pixel confidence level.

[0115] The trace probability evaluation index is obtained by the following formula:

[0116]

[0117] Where H(x) m y m ) represents the target pixel (x) m y m The corresponding trace probability evaluation index; R(x) m y m ) represents the target pixel (x) m y m The corresponding resonance score value; F(x) m y m ) represents the target pixel (x) m y m The confidence score of the corresponding pixel, where the target pixel (x) m y m The confidence score of the corresponding pixel can be obtained through the Bayesian confidence algorithm.

[0118] The trace probability evaluation index corresponding to the target pixel is compared with a preset index threshold; wherein the preset index threshold ranges from 0.42 to 0.51.

[0119] Target pixels whose trace probability evaluation index is not lower than a preset index threshold are taken as valid trace pixels, and the valid trace pixels are integrated to form a dotted distribution area.

[0120] In this embodiment, resonance scoring combined with confidence level filters pixels from two dimensions: physical characteristics and statistical reliability. Faced with complex grayscale distributions on sample surfaces due to material heterogeneity (e.g., mixing of different components, differences in surface oxide layers) and scanning noise (uniform electron beam scattering), a single grayscale threshold / feature is prone to misjudgment, while multi-dimensional fusion can more accurately distinguish between "real traces" and "interference noise." For example, on composite material surfaces, the grayscale differences in electron micrographs of different components are small, but their resonance characteristics (due to differences in atomic scattering and electronic transition characteristics) differ. Combining confidence level can accurately identify trace points of specific components. Furthermore, the physical mechanisms of surface traces differ among different samples (metals, semiconductors, biomaterials, etc.) (e.g., electron scattering patterns of metal scratches, resonance responses of protein aggregation on biofilm surfaces). However, the "resonance scoring + confidence level" framework in this solution does not depend on the grayscale patterns of specific samples and can be adapted to different scenarios by adjusting the "resonance scoring algorithm details" and "confidence level prior distribution." Compared to traditional methods that rely on fixed grayscale models, this approach is more versatile and expands the applicability of the technical solution. Meanwhile, by combining resonance evaluation and confidence level as described in this embodiment, even with the index threshold value reduced to 0.42-0.51, and under the condition of maximizing the effectiveness constraint of target pixels, it is still possible to obtain effective pixels without noise interference. This maximizes the accuracy of effective pixel selection and reduces the missed selection rate without interference. Furthermore, the trace probability evaluation index integrates "resonance intensity (R)" and "confidence level (F)," going beyond a binary judgment of "whether it is a valid trace" to quantify the "degree of effectiveness" of the trace, further improving the accuracy of trace pixel validity determination.

[0121] Specifically, resonance scoring is performed on candidate pixels to obtain the resonance score value corresponding to each candidate pixel, including:

[0122] The preprocessed electron microscopy image is subjected to multi-scale feature extraction to obtain multi-scale features; wherein, the multi-scale features include spatial scale features, material intrinsic features, geometric topological features and dynamic process features;

[0123] Local features of each candidate pixel are extracted, wherein the local features of the candidate pixel include gray-level distribution, texture pattern and physical field parameters;

[0124] Determine the cosine similarity S(x, y) between the feature vector formed by the local features of each candidate pixel and the feature vector formed by the multi-scale features.

[0125] Calculating the phase gradient Φ from electron micrographs BSE ;

[0126] A topology network is established on the candidate pixels, and a topology continuity score is calculated for each candidate pixel in the topology network.

[0127] The resonance score corresponding to candidate pixel (x, y) is obtained by combining the topological continuity score of each candidate pixel with the geometric constraint factor of each candidate pixel.

[0128] The resonance score value is obtained by the following formula:

[0129]

[0130] Where R(x,y) represents the resonance score value corresponding to the candidate pixel (x,y); S(x,y) represents the cosine similarity between the feature vector formed by the local features of the candidate pixel (x,y) and the feature vector formed by the multi-scale features. The gradient magnitude of the intrinsic phase field of the material is represented by S(x,y). The intrinsic phase field Φ of the material is extracted from the BSE image by Fourier phase decoupling technology, and then the magnitude of its gradient (i.e., the degree of drastic phase change) is calculated. Y(x,y) represents the geometric constraint factor corresponding to the candidate pixel (x,y), and its value range is (0,1]. Г represents the topological continuity coefficient. Specifically, S(x,y) is used to analyze the second phase particles of an alloy. If a pixel is the edge of the second phase particle, its local grayscale, texture (local features), and global multi-scale features (spatial distribution, composition contrast, and geometric morphology of the second phase) have high cosine similarity, indicating that it "belongs to a part of the second phase structure". This is the basic judgment for identifying phase boundaries and phase distribution. This is used to couple "physical field changes (phase gradient)" and "geometric constraints (Y)" to suppress noise caused by drastic changes in the physical field but irregular geometric shapes, and to enhance realistic structures with reasonable physical field changes and matching geometric shapes. When analyzing composite materials, if a region has a large phase gradient (suspected phase interface) but a low geometric constraint factor (disordered shape, unlike a real interface), the exponential term will weaken the score of that point to avoid misjudging it as a "valid structure"; conversely, a real phase interface (large phase gradient + regular geometric shape) will be enhanced.

[0131] The topological continuity coefficient is obtained by the following formula:

[0132]

[0133] Where ρ represents the topological continuity score of the target pixel obtained based on Morse-Small complex and persistence analysis; d represents the Riemannian manifold distance.

[0134] In this embodiment, existing technologies for extracting features from electron microscopy images often focus on a single scale (e.g., only microscopic morphology) or a single type (e.g., only compositional contrast). This embodiment proposes multi-scale features including spatial scale, material intrinsics, geometric topology, and dynamic processes, comprehensively covering the hierarchical features of material microstructure from the dimensions of "static structure-dynamic evolution" and "macroscopic distribution-microscopic properties." For example, when analyzing nanocomposite materials, it can extract not only the geometric morphology of nanoparticles (geometric topology) but also their spatial distribution at the micrometer scale (spatial scale) and the intrinsic properties reflected by compositional contrast (material intrinsics), resulting in a more comprehensive feature characterization. It also supplements local features such as grayscale distribution, texture patterns, and physical field parameters, achieving "microscopic-refined" feature supplementation for candidate pixels. Existing technologies for pixel analysis often remain at the grayscale or simple texture level. This embodiment introduces physical field parameters (e.g., atomic number fields associated with BSE images) to more closely bind pixel features to the physical essence of the material (e.g., composition, phase state), providing a more accurate foundational input for subsequent resonance scoring. Existing technologies utilize BSE images, directly mapping composition / phase distribution through multi-focus grayscale contrast. This embodiment's solution extracts the material's intrinsic phase field through Fourier phase decoupling and then calculates the phase gradient, upgrading traditional "static contrast analysis" to "dynamic phase change analysis." For example, when analyzing phase transition interfaces, the phase gradient can accurately characterize the intensity of atomic arrangement and compositional transitions at the interface, better reflecting the material's physical essence than simple grayscale differences and enhancing the physical correlation of features. Morse-Small complex and persistence analysis are introduced to construct a topological network, supplementing features from the dimension of "topological continuity." Existing technologies for pixel correlation analysis often rely on simple spatial distance or grayscale similarity. This embodiment's solution incorporates pixels into "structural network" analysis (such as pore connectivity and the topological morphology of phase distribution) through topological continuity scoring. For example, when analyzing porous materials, it can accurately identify continuous and isolated regions of the pore network, injecting physical meaning from the topological dimension into the resonance scoring. Existing pixel scoring technologies (such as saliency scoring and phase recognition scoring) often rely on single features (such as grayscale matching and shape matching). This embodiment couples cosine similarity (feature matching degree), phase gradient (physical field change), geometric constraint factor (spatial morphology), and topological continuity coefficient (structural correlation) to comprehensively evaluate the "resonance characteristics" of pixels from four dimensions: "feature matching - physical change - spatial morphology - topological correlation." For example, in heterogeneous interface recognition, it considers not only the cosine similarity of the features of the components on both sides but also the phase gradient at the interface (reflecting the intensity of component transition), geometric constraints (whether the interface morphology is regular), and topological continuity (whether the interface is part of a continuous structure), making the scoring more closely reflect the complex reality of material microstructures. The topological continuity coefficient is calculated based on Riemannian manifold distance, extending the analysis of pixel topological features from "discrete space" to "manifold space."Existing technologies for quantifying topological features mostly rely on counting simple connected components. The technical solution described in this embodiment utilizes Riemannian manifold distance to more accurately characterize the continuity of topological structures in "curved / non-Euclidean space" (such as the topological morphology of complex phase interfaces), thereby improving the quantification accuracy of topological features.

[0135] Compared to existing technologies, the technical solution described in this embodiment addresses the problem of "incomplete feature characterization" through multi-scale and multi-type feature coverage, the problem of "weak physical-topological correlation" through phase gradient and topological network, and the problem of "poor adaptability to complex microstructures" through multi-factor coupled scoring. Ultimately, this achieves more accurate and material-physical-intrinsic resonance scoring of electron micrograph pixels, significantly improving the accuracy and physical meaning of microstructure feature extraction and analysis, especially in scenarios such as heterogeneous interface identification, complex multiphase structure analysis, and dynamic evolution process tracking.

[0136] To address the problem in existing technologies where there is no specific processing and protection for flexible films, leading to easy deformation of the film surface, please refer to [link to relevant documentation]. Figure 1 This embodiment provides the following technical solution:

[0137] After preliminary protection, the sample undergoes rough processing, followed by sample separation, including:

[0138] Before roughing, the parameters of the focused ion beam electron beam dual-beam electron microscope are set, including adjusting the ion beam angle of the focused ion beam electron beam dual-beam electron microscope to 52°, the accelerating voltage to 30kV, and the ion beam current to 9.3nA.

[0139] Centered on the sample that has been initially protected, the upper and lower sides of the target area are milled using an ion beam, and the surrounding material of the target area is hollowed out, forming a preliminary thin sheet outline after hollowing out.

[0140] After forming the initial thin-film outline, the ion beam angle is changed to 0°, the voltage is maintained at 30kV, the current is reduced to 2.5nA, and U-shaped milling is performed on the bottom and sides of the target area.

[0141] After U-shaped milling, the ion beam angle is restored to 52°, the voltage is 30kV, and the current is 2.5nA to clean up the splashed material generated during the U-shaped milling process;

[0142] After cleaning, the sample undergoes rough processing, followed by sample separation.

[0143] Sample separation involves first switching the focused ion beam electron beam dual-beam microscope to low current mode, with an ion beam angle of 0° and a voltage of 30kV. The micromanipulation needle is then lowered above the target area of ​​the rough-processed sample, so that the tip of the micromanipulation needle contacts the surface of the target area. Carbon is deposited by ion beam induction to adhere and fix the micromanipulation needle to the target area.

[0144] After adhesion, lift the micromanipulation needle upwards to detach it from the substrate, then rotate it 180° so that the tip of the micromanipulation needle is pointing downwards and the bottom of the target area is facing upwards. Place the rotated sample needle tip into the sample stage, connect the sample to the sample stage using carbon, and then sever the connection between the micromanipulation needle and the sample.

[0145] Rotate the sample stage 180° again so that the bottom of the target area is facing upwards. Lower the micromanipulation needle again to contact the target area. After connecting with carbon, disconnect the sample from the sample stage and extract the sample back onto the micromanipulation needle. Rotate the needle 180° again to restore the sample to its normal orientation.

[0146] Finally, move the micromanipulation needle above the copper grid, adjust its position so that the target area is aligned with the copper grid mesh, use carbon to bond the sample to the side of the copper grid, then use an ion beam to cut the connection between the sample and the micromanipulation needle, pull out the needle, and complete the fixation of the sample on the copper grid.

[0147] The separation of the samples was finally completed.

[0148] Specifically, precise parameter settings are performed on the focused ion beam-electron beam dual-beam electron microscope (FEM) before rough machining. Setting specific values ​​for the ion beam angle, accelerating voltage, and current ensures the ion beam acts on the sample with appropriate energy and direction, avoiding over- or under-machining due to improper parameters. A step-by-step milling strategy is employed: first, the upper and lower sides of the target area are milled, and the surrounding material is removed to form a preliminary thin-film outline. Then, U-shaped milling and backsplashing material cleaning are performed. This gradual machining method effectively controls the sample morphology, reduces the damage to the sample structure caused by machining stress, and ensures the integrity and original properties of the sample. Step-by-step operation also facilitates timely detection and correction of deviations during machining, improving the success rate and sample qualification rate. During sample separation, ion beam-induced carbon deposition is used to achieve adhesion between the micromanipulation needle and the sample, as well as the connection between the sample and the sample stage and copper grid. This connection method is robust and causes minimal damage to the sample. By rotating the micromanipulation needle and sample stage multiple times, combined with precise adhesion and cutting operations, the sample can be kept stable during separation and transfer, avoiding sample drop or damage, and ultimately achieving precise fixation of the sample on the copper grid, providing a reliable guarantee for subsequent observation and analysis.

[0149] The separated samples undergo secondary protection, including:

[0150] The ion beam angle of the focused ion beam electron beam dual-beam electron microscope was adjusted to 0°, the accelerating voltage was set to 30kV, and the ion beam current was reduced to 80pA.

[0151] Initiate the ion beam-induced deposition process to deposit a 1 μm thick protective layer on the lower surface of the sample fixed on the copper grid using tungsten as the material;

[0152] During the deposition process, the ion beam continuously scans the bottom of the sample, ensuring that the tungsten material is uniformly covered and forming a continuous tungsten layer;

[0153] After deposition, the sample cross-section was observed in electron beam mode to confirm that the upper protective layer, sample layer and lower protective layer formed a complete three-layer structure. The upper protective layer is the tungsten layer deposited during the initial protection and the lower protective layer is the tungsten layer deposited during the secondary protection.

[0154] Finally, the secondary protection of the sample was completed.

[0155] Specifically, setting the ion beam angle to 0°, the accelerating voltage to 30kV, and the current to 80pA effectively balances processing efficiency and damage control. The 0° vertical angle ensures the ion beam energy acts perpendicularly on the sample surface, reducing lateral sputtering damage; the 30kV voltage meets the energy requirements for tungsten deposition while avoiding excessive energy that could cause film deformation or decomposition; the low 80pA current reduces the thermal effect and bombardment damage of the ion beam on the sample, making it particularly suitable for materials sensitive to temperature and mechanical stress, such as hydrophobic flexible films, ensuring the integrity of their chemical structure and micro / nano morphology. A 1μm protective layer is deposited on the lower surface of the sample using tungsten, utilizing its high hardness, high chemical stability, and good thermal conductivity to provide rigid support for the film. Continuous ion beam scanning of the sample bottom achieves uniform coverage, ensuring a tight fit between the protective layer and the sample surface, forming a continuous, gapless protective structure. This directional deposition method not only resists the mechanical stress during subsequent thinning and sample preparation, but also prevents the film from curling or breaking due to uneven stress during cross-sectional observation, maintaining its original micro-nano structural characteristics and providing a reliable foundation for subsequent imaging analysis. After deposition, the sample cross-section is observed in electron beam mode, and the integrity of the three-layer structure of "upper protective layer-sample layer-lower protective layer" is visually confirmed, establishing a closed-loop quality monitoring system. This visual verification method can promptly detect deposition defects (such as uneven thickness and interlayer bubbles), avoiding sample failure due to improper protection and ensuring that each prepared sample meets the requirements for high-precision characterization. Simultaneously, this process also provides feedback for subsequent sample preparation parameter adjustments, optimizing process stability and repeatability.

[0156] To address the issues in existing technologies where there is no targeted thickness removal of the flexible film and no effective measurement of the final flexible film, leading to inaccurate final film data, please refer to [link to relevant documentation]. Figure 1This embodiment provides the following technical solution:

[0157] The sample after secondary protection is thinned and prepared, including:

[0158] The thinned area of ​​the sample after secondary protection was confirmed;

[0159] After confirming the thinning region, the parameters of the focused ion beam and electron beam dual-beam electron microscope were adjusted. Specifically, the ion beam angle was set to 52°, and the accelerating voltage was divided into a coarse thinning stage and a fine thinning stage. The current in the coarse thinning stage was set to 0.43 nA, and the current in the fine thinning stage was reduced to 80 pA.

[0160] The thinning area is first subjected to coarse thinning, which is as follows: the thinning area is milled with an ion beam angle of 52° and a current of 0.43nA, and the material is removed layer by layer from the top protective layer to the bottom until the remaining thickness is 100nm. When the milling reaches the preset time, the SEM mode is switched to scan and the remaining thickness is measured.

[0161] The thinned area is then further thinned by reducing the ion beam current to 80 pA and milling the thinned area. The thickness of the sample is measured after each milling until the sample thickness reaches 50 nm.

[0162] After the sample is thinned, the sample surface is scanned with an ion beam. If local thickness inhomogeneity is found after scanning, an 80pA current is used to refine the thick area.

[0163] Then clean the surface of the sample that has been fixed to the copper grid and thinned.

[0164] Finally, the sample thinning and preparation were completed.

[0165] Specifically, through a clearly defined step design and parameter setting, a scientific and efficient operational path is provided for sample thinning preparation. First, the thinning area of ​​the sample after secondary protection is confirmed. This operation precisely identifies the part that needs to be processed, avoiding damage to other areas of the sample caused by blind thinning, ensuring sample integrity and the validity of experimental results. Simultaneously, based on area confirmation, subsequent parameter adjustments and thinning operations can be made specifically, improving the accuracy and efficiency of sample preparation. The ion beam angle is fixed at 52°, and different accelerating voltages and currents are set for the coarse and fine thinning stages. A current of 0.43 nA in the coarse thinning stage can quickly remove a large amount of material, shortening sample preparation time and improving processing efficiency; while the current is reduced to 80 pA in the fine thinning stage to achieve fine removal of material, ensuring the accuracy of the thinning thickness. This staged parameter setting balances sample preparation efficiency and accuracy, meeting the needs of different stages. The operation process of coarse thinning followed by fine thinning makes the sample preparation process more orderly and controllable. Coarse thinning quickly removes most of the material, reducing overall processing time. When approaching the target thickness, it switches to fine thinning, and combined with thickness monitoring after each milling operation, it can precisely control the final sample thickness to 50 nm. Subsequent finishing of localized uneven areas and surface cleaning further ensures sample quality, guaranteeing a smooth and uniform sample surface after thinning. This provides a high-quality sample foundation for subsequent analysis and significantly improves the reliability and accuracy of experimental data.

[0166] After thinning and sample preparation, microstructure imaging was performed under low-damage conditions to obtain film size and morphology data, including:

[0167] The thinned sample is loaded into the sample holder of the focused ion beam electron beam dual-beam electron microscope.

[0168] In a dual-beam focused ion beam electron beam microscope, the edge of the copper grid or a non-target area is first selected as the focal point. The sample is coarsely focused by low magnification. At the same time, the electron beam avoids the target area of ​​the sample during the focusing process.

[0169] Then, the target area of ​​the sample on the copper grid is located by low-magnification scanning. At the same time, the magnification is increased to 10000× to confirm the sample position.

[0170] Fine focusing is performed in a blank area far from the target area of ​​the sample. After focusing, the target area of ​​the sample is moved quickly by moving the sample stage to the center of the field of view.

[0171] After the sample is translated, a high-sensitivity CCD camera is used to expose the sample. If multiple areas are observed, the field of view is switched quickly in sequence, and each area is photographed no more than twice to obtain the final measurement image.

[0172] The thickness, surface undulation amplitude, and microstructure feature size of the thin film in the measurement image are measured using a ruler tool;

[0173] The final measurement data are characterization data for hydrophobic micro / nano-scale flexible films.

[0174] Specifically, coarse focusing is first performed on the edge of the copper grid or a non-target area, with the electron beam avoiding the target area to effectively prevent direct bombardment and damage to the sample. Fine focusing is then performed on the blank area, and the target area is moved to the center of the field of view by translating the sample stage. This step-by-step focusing strategy minimizes the physical impact on the target area of ​​the sample, ensuring that the sample remains in its original state during imaging and measurement, providing a foundation for accurate characterization. Using a focused ion beam electron beam dual-beam microscope, after locating the target area through low-magnification scanning, the magnification is increased to 10000×, clearly capturing the microstructural details of the sample. Combined with a high-sensitivity CCD camera for exposure imaging, if multiple areas need to be observed, the field of view can be quickly switched, with each area photographed no more than twice. This ensures imaging efficiency while avoiding potential damage to the sample due to overexposure. The acquired measurement images accurately reflect the sample morphology, providing a reliable basis for subsequent data measurement. Using a scale tool, the thickness, surface undulation amplitude, and microstructural feature size of the thin film in the measurement images can be measured, converting the morphological characteristics of the hydrophobic micro / nano-scale flexible thin film into precise data. This quantitative measurement method is more scientific and convincing than qualitative observation. The final measurement data obtained can serve as characterization data for thin films, providing key information for material performance research and process optimization, thus enhancing the value and practicality of the entire research.

[0175] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0176] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A characterization method for hydrophobic micro / nano-scale flexible thin film structures, applied to hydrophobic micro / nano-scale flexible thin film structures, characterized in that, Hydrophobic micro / nano-scale flexible thin film structures include flexible thin film body structure, surface protection structure, conductive structure, support structure and fixing structure; The surface protection structure is formed by lightly drawing on the thin film surface with an oil-based marker, creating a dotted ink distribution due to hydrophobicity, which is used to mark the sampling area and protect the surface; the conductive structure is a gold vapor deposition layer; the support structure includes an upper protective layer, a sample layer, and a lower protective layer. The fixed structure is a copper grid substrate; Characterization methods for hydrophobic micro / nano-scale flexible thin film structures include: First, the thin film sample is pre-processed. The pre-processed sample is marked with the target area and the marked target area is initially protected. The sample with initial protection is then rough processed. After rough processing, the sample is separated. The separated sample is then protected a second time. The sample with secondary protection is then thinned and prepared. After thinning and preparation, microstructure imaging is performed under low-damage conditions to obtain thin film size and morphology data. Sample pretreatment of the thin film sample includes: Draw a line on the surface of the hydrophobic film with an oil-based marker. After drawing the line, the film surface will be dotted and form a protective surface structure. Let the surface of the film stand until the ink dries; The dried film sample is placed in the gold vapor deposition equipment, and a gold film is vapor deposited on the surface of the film sample. Finally, the sample pretreatment of the thin film sample is completed; The pretreated sample is then marked with the target region, and the marked target region is initially protected, including: The pretreated thin film sample was placed in the sample chamber of a dual-beam electron microscope with a focused ion beam and an electron beam, and the sample surface was scanned at low magnification using an electron beam. Based on the scanning results, the dotted traces on the sample surface are identified; The identified point-like distribution area is taken as the target area, and the boundary of the target area is marked using the marking function of the focused ion beam electron beam dual-beam electron microscope. Next, the target area is initially protected by adjusting the ion beam angle of the focused ion beam electron beam dual-beam electron microscope, wherein the ion beam angle is 52°, the accelerating voltage is set to 30kV, and the ion beam current is set to 0.23nA. A tungsten protective layer with a thickness of 1 μm was deposited on the surface of the target area; The deposition process is initiated, and tungsten material covers the target area under the action of an ion beam, forming the initial structure of the upper protective layer and the sample layer.

2. The characterization method for hydrophobic micro / nano-scale flexible thin film structures according to claim 1, characterized in that, Based on the scanning results, the dotted distribution of traces on the sample surface is identified, including: The electron micrographs obtained by scanning the sample surface are preprocessed to obtain preprocessed electron micrographs; wherein, the preprocessing includes grayscale normalization and noise suppression. Retrieve the global grayscale mean and grayscale standard deviation of the preprocessed electron micrograph; A dynamic threshold is set using the global grayscale average value and grayscale standard deviation of the preprocessed electron micrograph; Each pixel in the electron micrograph is compared with a dynamic threshold, and pixels with gray values ​​exceeding the dynamic threshold are selected as candidate pixels. Resonance scoring is performed on candidate pixels to obtain the resonance score value corresponding to each candidate pixel; The resonance score value is compared with a preset score threshold, wherein the preset score threshold ranges from 0.62 to 0.

67. Candidate pixels whose resonance score values ​​exceed a preset score threshold are selected as target pixels. For each target pixel, a trace probability evaluation index is generated using the resonance score and the pixel confidence level. The trace probability evaluation index corresponding to the target pixel is compared with a preset index threshold; wherein the preset index threshold ranges from 0.42 to 0.

51. Target pixels whose trace probability evaluation index is not lower than a preset index threshold are taken as valid trace pixels, and the valid trace pixels are integrated to form a dotted distribution area.

3. The characterization method for hydrophobic micro / nano-scale flexible thin film structures according to claim 2, characterized in that, Resonance scoring is performed on candidate pixels to obtain the resonance score value for each candidate pixel, including: The preprocessed electron microscopy image is subjected to multi-scale feature extraction to obtain multi-scale features; wherein, the multi-scale features include spatial scale features, material intrinsic features, geometric topological features and dynamic process features; Local features of each candidate pixel are extracted, wherein the local features of the candidate pixel include gray-level distribution, texture pattern and physical field parameters; Determine the cosine similarity S(x, y) between the feature vector formed by the local features of each candidate pixel and the feature vector formed by the multi-scale features. Calculating the phase gradient Φ from electron micrographs BSE ; A topology network is established on the candidate pixels, and a topology continuity score is calculated for each candidate pixel in the topology network. The resonance score corresponding to candidate pixel (x, y) is obtained by combining the topological continuity score of each candidate pixel with the geometric constraint factor of each candidate pixel.

4. The characterization method for hydrophobic micro / nano-scale flexible thin film structures according to claim 3, characterized in that, After preliminary protection, the sample undergoes rough processing, followed by sample separation, including: Before roughing, the parameters of the focused ion beam electron beam dual-beam electron microscope are set, including adjusting the ion beam angle of the focused ion beam electron beam dual-beam electron microscope to 52°, the accelerating voltage to 30kV, and the ion beam current to 9.3nA. Centered on the sample that has been initially protected, the upper and lower sides of the target area are milled using an ion beam, and the surrounding material of the target area is hollowed out, forming a preliminary thin sheet outline after hollowing out. After forming the initial thin-film outline, the ion beam angle is changed to 0°, the voltage is maintained at 30kV, the current is reduced to 2.5nA, and U-shaped milling is performed on the bottom and sides of the target area. After U-shaped milling, the ion beam angle is restored to 52°, the voltage is 30kV, and the current is 2.5nA to clean up the splashed material generated during the U-shaped milling process; After cleaning, the sample undergoes rough processing, followed by sample separation. Sample separation involves first switching the focused ion beam electron beam dual-beam microscope to low current mode, with an ion beam angle of 0° and a voltage of 30kV. The micromanipulation needle is then lowered above the target area of ​​the rough-processed sample, so that the tip of the micromanipulation needle contacts the surface of the target area. Carbon is deposited by ion beam induction to adhere and fix the micromanipulation needle to the target area. After adhesion, lift the micromanipulation needle upwards to detach it from the substrate, then rotate it 180° so that the tip of the micromanipulation needle is pointing downwards and the bottom of the target area is facing upwards. Place the rotated sample needle tip into the sample stage, connect the sample to the sample stage using carbon, and then sever the connection between the micromanipulation needle and the sample. Rotate the sample stage 180° again so that the bottom of the target area is facing upwards. Lower the micromanipulation needle again to contact the target area. After connecting with carbon, disconnect the sample from the sample stage and extract the sample back onto the micromanipulation needle. Rotate the needle 180° again to restore the sample to its normal orientation. Finally, move the micromanipulation needle above the copper grid, adjust its position so that the target area is aligned with the copper grid mesh, use carbon to bond the sample to the side of the copper grid, then use an ion beam to cut the connection between the sample and the micromanipulation needle, pull out the needle, and complete the fixation of the sample on the copper grid. The separation of the samples was finally completed.

5. The characterization method for hydrophobic micro / nano-scale flexible thin film structures according to claim 4, characterized in that, The separated samples undergo secondary protection, including: The ion beam angle of the focused ion beam electron beam dual-beam electron microscope was adjusted to 0°, the accelerating voltage was set to 30kV, and the ion beam current was reduced to 80pA. Initiate the ion beam-induced deposition process to deposit a 1 μm thick protective layer on the lower surface of the sample fixed on the copper grid using tungsten as the material; During the deposition process, the ion beam continuously scans the bottom of the sample, ensuring uniform coverage of the tungsten material and forming a continuous tungsten layer; After deposition, the sample cross-section was observed in electron beam mode to confirm that the upper protective layer, sample layer and lower protective layer formed a complete three-layer structure. The upper protective layer is the tungsten layer deposited during the initial protection and the lower protective layer is the tungsten layer deposited during the secondary protection. Finally, the secondary protection of the sample was completed.

6. The characterization method for hydrophobic micro / nano-scale flexible thin film structures according to claim 5, characterized in that, The sample after secondary protection is thinned and prepared, including: The thinned area of ​​the sample after secondary protection was confirmed; After confirming the thinning region, the parameters of the focused ion beam and electron beam dual-beam electron microscope were adjusted. Specifically, the ion beam angle was set to 52°, and the accelerating voltage was divided into a coarse thinning stage and a fine thinning stage. The current in the coarse thinning stage was set to 0.43 nA, and the current in the fine thinning stage was reduced to 80 pA. The thinning area is first subjected to coarse thinning, which is as follows: the thinning area is milled with an ion beam angle of 52° and a current of 0.43nA, and the material is removed layer by layer from the top protective layer to the bottom until the remaining thickness is 100nm. When the milling reaches the preset time, the SEM mode is switched to scan and the remaining thickness is measured. The thinned area is then further thinned by reducing the ion beam current to 80 pA and milling the thinned area. The thickness of the sample is measured after each milling until the sample thickness reaches 50 nm. After the sample is thinned, the sample surface is scanned with an ion beam. If local thickness inhomogeneity is found after scanning, an 80pA current is used to refine the thick area. Then clean the surface of the sample that has been fixed to the copper grid and thinned. Finally, the sample thinning and preparation were completed.

7. The characterization method for hydrophobic micro / nano-scale flexible thin film structures according to claim 6, characterized in that, After thinning and sample preparation, microstructure imaging was performed under low-damage conditions to obtain film size and morphology data, including: The thinned sample is loaded into the sample holder of the focused ion beam electron beam dual-beam electron microscope. In a dual-beam focused ion beam electron beam microscope, the edge of the copper grid or a non-target area is first selected as the focal point. The sample is coarsely focused by low magnification. At the same time, the electron beam avoids the target area of ​​the sample during the focusing process. Then, the target area of ​​the sample on the copper grid is located by low-magnification scanning. At the same time, the magnification is increased to 10000× to confirm the sample position. Fine focusing is performed in a blank area far from the target area of ​​the sample. After focusing, the target area of ​​the sample is moved quickly by moving the sample stage to the center of the field of view. After the sample is translated, a high-sensitivity CCD camera is used to expose the sample. If multiple areas are observed, the field of view is switched quickly in sequence, and each area is photographed no more than twice to obtain the final measurement image. The thickness, surface undulation amplitude, and microstructure feature size of the thin film in the measurement image are measured using a ruler tool; The final measurement data are characterization data for hydrophobic micro / nano-scale flexible films.

Images