A method for detecting the conductive performance of a conductive film nanosilver coating

Abstract

The present application belongs to the technical field of conductive film nanometer silver coating conductive performance detection, and specifically discloses a conductive film nanometer silver coating conductive performance detection method, which divides the nanometer silver coating into regular grid units, synchronously extracts the topographic features and electrical features of each unit, and verifies the trend based on the global statistical correlation of the two, if the global statistical correlation is established, anisotropy consistency defect detection is carried out, otherwise, global multidimensional anomaly screening is started, the strategy realizes the deep diagnosis of the intrinsic factor of the coupling relationship of the conductive network structure performance, significantly improves the guiding value of the detection result for process optimization, under the premise of judging the anomaly based on the statistical correlation of the topographic features and electrical features of the grid units, the abnormal grid is identified through the comprehensive analysis of the grid unit level line density, average square resistance and its consistency deviation or extreme value distribution, and the spatial fine mapping and classified characterization of the defect distribution are realized.

Description

A method for detecting the conductivity of a conductive film nano-silver coating Technical Field

[0001] This invention belongs to the field of conductive film nano-silver coating conductivity testing technology, and specifically discloses a method for testing the conductivity of conductive film nano-silver coating. Background Technology

[0002] Transparent conductive films made of nano-silver are widely used in touchscreens, flexible displays, and solar cells due to their excellent photoelectric properties, flexibility, and solution processability. As the core functional layer of the conductive film, the conductivity of the nano-silver coating directly determines the reliability and performance ceiling of the final device. Therefore, quality testing of the conductivity of the nano-silver coating is a crucial step in ensuring product consistency.

[0003] Existing technologies already include technical solutions for detecting the conductivity of nano-silver coatings. For example, Chinese Invention Patent Publication No. CN118501605B discloses a quality detection system for nano-silver transparent conductive films based on machine vision. This system decomposes the performance evaluation of samples into three parallel dimensions: conductivity, transparency, and flatness. In the conductivity detection module, multiple monitoring points are set up on the surface of a single sample to measure the sheet resistance, thereby calculating the sheet resistance uniformity and sheet resistance compliance rate. The system also integrates surface dirt information obtained from image recognition and finally outputs a comprehensive conductivity index for quality assessment.

[0004] The above scheme describes the fluctuation of sheet resistance on the sample surface by calculating statistical quantities such as sheet resistance uniformity. However, it only reflects the macroscopic distribution characteristics of electrical performance and remains at the level of phenomenon description. It cannot reveal the intrinsic causes of conductivity fluctuations. Furthermore, when evaluating conductivity, external interference factors such as surface dirt are taken into consideration as influencing factors. The influence of dirt can be eliminated by cleaning, but the microstructure of the silver nanowire network is the fundamental factor that determines the upper limit of the coating's conductivity. Ignoring the in-depth diagnosis of intrinsic factors weakens the guiding value of the test results for process improvement to a certain extent.

[0005] Furthermore, the above solution ultimately outputs a comprehensive score for conductivity performance. This highly generalized result lacks spatial resolution. When test results indicate quality abnormalities, production personnel struggle to pinpoint the exact problem based on a generalized score, thus prolonging the process debugging cycle. Summary of the Invention

[0006] To solve the above-mentioned technical problems, or at least partially solve them, the present invention provides a method for detecting the conductivity of a conductive film nano-silver coating.

[0007] The objective of this invention can be achieved through the following technical solution: a method for detecting the conductivity of a conductive film nano-silver coating, comprising the following steps: establishing a rectangular grid array on the surface of the nano-silver coating.

[0008] Within each grid cell of the rectangular grid array, images of the microstructure of silver nanowires and data on their transverse and longitudinal sheet resistance were collected.

[0009] Based on the microstructure image of each grid cell and the sheet resistance data in the lateral and longitudinal directions, the linear density, dominant orientation angle, average sheet resistance, and low resistance direction reflecting the electrical characteristics of the silver nanowires are extracted.

[0010] Global trend verification was performed on all grid cells by analyzing the statistical correlation between line density and average sheet resistance.

[0011] If the global trend verification passes, the anisotropy verification will proceed. By comparing the dominant orientation angle of the silver nanowires with the low resistance direction in each grid cell, hidden anomalous grids will be identified.

[0012] If the global trend verification fails, the process proceeds to the global anomaly screening stage. By analyzing the regular deviations or independent anomalies between the line density and average sheet resistance of each grid cell, explicit anomalous grids are identified.

[0013] Spatial clustering analysis is performed on all abnormal meshes to aggregate them into defect regions with different attributes.

[0014] Combining all the above technical solutions, the positive effects of this invention are as follows: 1. This invention divides the nano-silver coating into regular grid units, simultaneously extracts the morphological and electrical features of each unit, and performs trend verification based on the global statistical correlation between the two. If the global statistical correlation is valid, anisotropic consistency defect detection is carried out; otherwise, a multi-dimensional anomaly screening is initiated. This strategy achieves in-depth diagnosis of the intrinsic factor of the coupling relationship between the performance of the conductive network structure, significantly improving the guiding value of the detection results for process optimization.

[0015] 2. Based on the statistical correlation between the morphological and electrical characteristics of grid cells to determine anomalies, this invention identifies abnormal grids by comprehensively analyzing the linear density, average sheet resistance, and consistency deviation or extreme value distribution of grid cells. This enables a refined spatial mapping and classification of defect distribution, which not only distinguishes different failure mechanisms but also significantly improves the accuracy and interpretability of anomaly detection. Attached Figure Description

[0016] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0017] Figure 1 is a diagram illustrating the implementation steps of the method of the present invention.

[0018] Figure 2 is a flowchart of the implementation of global trend verification in this invention.

[0019] Figure 3 is a flowchart of the implementation of the full-domain anomaly screening corresponding to the regular deviation anomaly in this invention. Detailed Implementation

[0020] 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.

[0021] Referring to Figure 1, the present invention proposes a method for detecting the conductivity of a conductive film nano-silver coating, comprising the following steps: S1, establishing a rectangular grid array on the surface of the nano-silver coating.

[0022] Given that the conductivity of the nano-silver coating is highly dependent on the local microstructure of the silver wire network, by dividing the continuous surface into regular rectangular units, the macro-coating can be decomposed into several micro-regions with clear spatial coordinates, thereby independently extracting morphology and electrical parameters within each unit and achieving localized and refined characterization.

[0023] In addition, the rectangular grid provides a unified spatial reference framework for different detection methods such as topography image acquisition and sheet resistance measurement, ensuring that the topography and electrical data of the same physical area strictly correspond and avoiding misjudgment due to positional offset.

[0024] Specifically, the process of establishing the rectangular grid array is as follows: with the geometric center of the nano-silver coating as the origin, a two-dimensional rectangular coordinate system is established, consisting of mutually orthogonal horizontal reference axes and vertical reference axes.

[0025] The establishment of the aforementioned two-dimensional rectangular coordinate system is essentially based on a two-dimensional Cartesian coordinate system. This coordinate system has the characteristics of orthogonal orientation and reversible spatial mapping. Specifically, the horizontal and vertical divisions are not coupled, and any grid can be uniquely deduced from its physical location through its row and column indices, supporting precise positioning.

[0026] Along the horizontal and vertical reference axes, the area is divided at equal intervals with a set spacing to generate a set of non-overlapping and seamless grid cells, forming a rectangular grid array that covers the area to be tested of the coating.

[0027] Each grid cell is assigned a spatial identifier, which contains its sequential position information on the horizontal and vertical reference axes, for spatial positioning of the grid cell.

[0028] S2. Within each grid cell of the rectangular grid array, collect images of the microstructure of the silver nanowires and data on their transverse and longitudinal sheet resistance.

[0029] The conductivity of a nano-silver coating does not primarily depend on the conductivity of the silver material itself, but rather on the network structure formed by the silver wires within the coating. This includes factors such as the quantity of silver wires, their connectivity, and their orientation. This microstructural information is not visible through ordinary electrical measurements and must be observed and analyzed using microscopic imaging.

[0030] On the other hand, sheet resistance, or resistance per unit area, is a commonly used indicator to measure the conductivity of thin films; the lower the sheet resistance, the better the conductivity. In actual production, such as during coating or scraping processes, silver wires often align more densely along a certain direction, resulting in better conductivity in one direction and poorer conductivity in another. Measuring only the sheet resistance in one direction may overlook this directional difference, leading to a misjudgment of the overall performance. Therefore, simultaneously measuring the sheet resistance in both the transverse and longitudinal directions provides a more comprehensive reflection of the coating's true conductivity.

[0031] Based on the above considerations, the following data were collected for the microstructure of silver nanowires and the transverse and longitudinal sheet resistance data in each grid cell: a composite detection device integrating a two-dimensional moving platform, optical imaging and a four-probe sheet resistance measurement device was used.

[0032] Understandably, in order to accurately determine the relationship between the microstructure and conductivity of the nano-silver coating, it is essential to ensure that the captured images and measured resistance data originate from the same location. Therefore, this invention first captures images within each grid cell and then measures the electrical properties, both concentrated at the cell center.

[0033] By using a high-precision two-dimensional moving platform, and integrating optical imaging and a four-probe sheet resistance measurement device on it, imaging and electrical measurements can be completed sequentially after accurately locating the center of a grid. In this way, the image and resistance data truly correspond to the same tiny area, achieving precise spatial matching of two different types of data and effectively avoiding erroneous judgments caused by inconsistent measurement positions.

[0034] A two-dimensional moving platform is used to drive the optical imaging device to move sequentially to the center area of ​​each grid cell and acquire the surface image of that cell as a microscopic morphology image.

[0035] In one specific embodiment, the optical imaging device can be a microscope camera.

[0036] Switch to the four-probe sheet resistance measurement device in the same moving position and control the four-probe assembly to contact the center point of the grid cell.

[0037] At the center point, measurement currents are applied sequentially along the predefined transverse and longitudinal measurement directions, and voltage signals are acquired. The transverse and longitudinal sheet resistance values ​​of the grid cell are then calculated.

[0038] It should be noted that the basic principle of the four-probe sheet resistance measurement technique is as follows: a constant current is injected through the two outer probes, and the voltage drop generated by this is detected by the two inner probes. Since the current in the voltage measurement circuit is extremely small, the contact resistance between the probes and the sample has a negligible impact on the measurement results. Based on this, and combined with the standard four-probe formula, the sheet resistance of the material in the transverse and longitudinal directions can be calculated separately.

[0039] It is worth noting that the raw data obtained from the transverse and longitudinal measurements are the applied current and the corresponding voltage response, from which the apparent resistance value can be calculated. However, this resistance value is affected by the probe arrangement, sample geometry, and boundary conditions, and cannot directly reflect the intrinsic conductivity of the material. Therefore, the standard four-probe formula needs to be used for conversion.

[0040] The standard four-probe sheet resistance conversion formula is based on the analysis of a uniform, semi-infinite thin-film conductor, establishing a deterministic relationship between apparent resistance and sheet resistance:

[0041] For a common linear four-probe configuration, under ideal conditions where the sample size is much larger than the probe spacing and edge effects are ignored, the sheet resistance can be calculated using the following formula: ,in Indicates obstruction. , These represent the voltage and current values, respectively. C represents the geometry correction factor; for a standard isotropic four-probe array, C≈4.532.

[0042] In actual testing, the apparent resistance in the two directions can be converted into the corresponding directional sheet resistance by substituting the current and voltage measured in the horizontal and vertical directions into the above formula.

[0043] S3. Based on the micro-morphology image of each grid cell and the transverse and longitudinal sheet resistance data, extract the linear density, dominant orientation angle, average sheet resistance, and low resistance direction that reflect the morphological characteristics of the silver nanowires.

[0044] After completing the microscopic morphology image and lateral and longitudinal sheet resistance data acquisition for each grid cell, it is necessary to further extract features to establish a quantitative correlation between the distribution of silver nanowires and their conductivity, thereby achieving accurate identification of abnormal areas.

[0045] In this process, the extracted morphological features include line density and dominant orientation angle.

[0046] Line density reflects the overall coverage level of the conductive network. The denser the silver lines, the higher the probability of forming a continuous conductive path, which usually corresponds to better overall conductivity.

[0047] The dominant orientation angle quantifies the orientation preference of silver lines in local areas by statistically analyzing the orientation distribution of skeleton pixels, reflecting the microstructural details of the distribution of nano-silver. Especially in processes such as coating and scraping, the preferential orientation of silver lines is often caused by fluid shearing, which in turn leads to electrical anisotropy.

[0048] Accordingly, the extracted electrical characteristics include average sheet resistance and low resistance direction.

[0049] The average sheet resistance represents the overall conductivity of the grid cells. It eliminates the randomness of measurements in a single direction and provides a stable and comparable electrical performance benchmark.

[0050] The low resistance direction reflects the optimal conduction orientation of conductivity in space and is a detailed indicator of the anisotropic characteristics of local electrical behavior.

[0051] Therefore, this invention simultaneously extracts overall and detailed indicators in both morphology and electrical dimensions, constructing a multi-scale feature system, which provides a data foundation for establishing the correlation between the distribution and conductivity of silver nanowires.

[0052] In a preferred embodiment of the present invention, the linear density and dominant orientation angle reflecting the morphological characteristics of the silver nanowires are extracted as follows: the micromorphological image of each grid cell is processed by image segmentation to separate the silver nanowire region in the image from the background region.

[0053] Understandably, silver nanowires typically appear as bright structures in microscopic images, exhibiting a significant difference in grayscale or color compared to the background. Image segmentation converts the image into a binary mask, enabling the separation of the target structure from the background.

[0054] The segmented silver nanowire regions were subjected to skeletonization processing to obtain a single-pixel-width silver nanowire skeleton network image.

[0055] To be further understood, the core of skeletonizing the silver nanowire region lies in peeling away the boundary pixels layer by layer, thereby preserving the centerline of a single pixel width. This method can ensure that the topological connectivity, branching pattern and orientation information of the original structure remain unchanged, while eliminating the length calculation error that may be caused by changes in line width.

[0056] This process makes the calculated length a pure geometric measure of the density of the silver nanowire distribution, that is, a true reflection of the linear density.

[0057] The sum of the actual physical lengths of all skeleton pixels within a unit area of ​​a grid cell is calculated based on the image of the silver nanowire skeleton network, and is defined as the line density of the grid cell.

[0058] Statistical analysis was performed on the tangent directions of all skeleton pixels in the image of the silver nanowire skeleton network to construct a histogram of skeleton direction distribution.

[0059] In an optional embodiment, the tangent direction of the skeleton pixel can be estimated by local gradient calculation to characterize the local orientation of the silver nanowire skeleton at that location. Since the linear structure has geometric directional symmetry, i.e. 0° and 180° represent the same straight line direction, the direction is usually defined and statistically analyzed with a period of 0° to 180°.

[0060] The median of the direction interval with the highest distribution frequency in the histogram of the skeleton orientation distribution is determined as the dominant orientation angle of the silver nanowire arrangement within that grid cell.

[0061] Since the direction interval with the highest frequency in the skeleton orientation distribution histogram corresponds to the direction in which the silver nanowires are most concentrated or predominantly arranged, selecting the median of this peak interval as the dominant orientation angle can effectively suppress the interference of local noise or outliers on orientation estimation and improve the accuracy of orientation characterization.

[0062] In another preferred embodiment of the present invention, the specific extraction process of the average sheet resistance and low resistance direction reflecting electrical characteristics is as follows: the lateral sheet resistance value and the longitudinal sheet resistance value measured in each grid cell are arithmetically averaged to obtain the average sheet resistance reflecting the overall conductivity level of the cell.

[0063] By comparing the lateral and longitudinal square resistance values ​​within each grid cell, the direction with the smaller value is determined as the low resistance direction of that grid cell.

[0064] S4. Verify the global trend of all grid cells by analyzing the statistical correlation between line density and average sheet resistance.

[0065] Considering that the nano-silver coating is not a complete metal film, but a network of silver wires that conduct electricity, the more silver wires there are, the higher the line density, the denser the overlap points, the more complete the conductive path, and the lower the average sheet resistance. Therefore, under normal process conditions, the line density and the average sheet resistance should show a significant negative correlation.

[0066] This invention, by first verifying the existence of this global trend, can quickly determine whether the overall conductivity of the coating is normal. Compared to simply looking at the sheet resistance value, this method can initially identify anomalies globally, providing a reliable basis and direction for subsequent, more detailed analysis of local defects.

[0067] Referring to Figure 2, specifically, the global trend verification process is as follows: the line densities of all grid cells are summarized to form a full-coating line density sequence.

[0068] The average sheet resistance of all grid cells is summed to form the average sheet resistance sequence of the entire coating.

[0069] Calculate the rank correlation coefficient between the full coating linear density parameter sequence and the full coating average sheet resistance sequence.

[0070] It should be noted that the rank correlation coefficient is essentially the conversion of linear density and average sheet resistance into sorted sequences, and the calculation of the Pearson correlation coefficient between these two rank sequences. It measures the consistency of the sorting order of the two variables. When the rank correlation coefficient approaches -1, it indicates that the two have a strong negative monotonic correlation, that is, the higher the linear density, the lower the average sheet resistance. This reflects that the conductive network structure and electrical properties in the nano-silver coating conform to the basic physical law that the sheet resistance decreases as the silver linear density increases.

[0071] The calculated rank correlation coefficient is compared with the strong negative correlation threshold;

[0072] If the rank correlation coefficient is less than or equal to the strong negative correlation threshold, the overall conductive network of the coating is determined to conform to the basic physical law that the sheet resistance decreases as the silver line density increases, and the global trend verification is passed. Otherwise, the overall conductive network of the coating is determined to not conform to this basic physical law, and the global trend verification is not passed.

[0073] The above-mentioned threshold for strong negative correlation can be set based on statistical experience of normal process samples. Specifically, firstly, by testing a large number of known good nano-silver coating samples, the rank correlation coefficient distribution between their linear density and average sheet resistance is calculated.

[0074] Secondly, take the lower limit of the distribution, such as the 5th percentile, or set a conservative value acceptable in engineering, such as -0.6, as the strong negative correlation threshold, which represents the minimum negative correlation intensity that most samples should reach under normal process conditions.

[0075] Since the smaller the rank correlation coefficient is when it is negative, the closer it is to -1, indicating a stronger negative correlation between linear density and average sheet resistance; therefore, when the rank correlation coefficient is less than or equal to the strong negative correlation threshold, the global trend verification is considered successful.

[0076] S5. If the global trend verification passes, proceed to the anisotropy verification. By comparing the dominant orientation angle of the silver nanowires with the low resistance direction in each grid cell, hidden anomalous grids are identified.

[0077] When the global trend verification is successful, it only indicates that the coating as a whole conforms to the macroscopic physical law that higher line density corresponds to lower average sheet resistance. It does not rule out the existence of hidden structural defects in localized areas. For example, the average sheet resistance of some areas may be within the normal range, but their sheet resistance anisotropy ratio may be significantly higher, reflecting a severe imbalance between lateral and longitudinal conductivity. Such anomalies usually originate from excessive concentration of silver wire orientation. Although they do not cause the overall resistance to exceed the standard, they may affect the reliability and uniformity of the device in actual use.

[0078] Such hidden structural defects are difficult to detect through global trends or a single average sheet resistance index, but anisotropy verification can effectively identify these hidden defects.

[0079] In one feasible approach to the above scheme, the anisotropy verification process is as follows:

[0080] Based on the low resistance direction of each grid cell, the corresponding theoretical conductivity advantage angle reference is determined; where, if the low resistance direction is transverse, the theoretical conductivity advantage angle reference is 0° or 180°; if the low resistance direction is longitudinal, the theoretical conductivity advantage angle reference is 90°.

[0081] Calculate the minimum angle difference between the dominant orientation angle of each grid cell and the theoretical conductivity advantage angle reference.

[0082] It should be added that, considering the periodicity and directional symmetry of angles within the range of 0° to 180°, the calculation of the minimum angle difference should follow the following principle: when the absolute value of the direct difference between two angles exceeds 90°, the supplementary angle should be taken as the angle difference.

[0083] The minimum angle difference is compared with the configured angle tolerance threshold, where the angle tolerance limit can refer to industry practice, and the common range is 10° to 20°.

[0084] If the minimum angle difference is greater than the angle tolerance threshold, the grid cell is determined to have a structural imbalance where the direction of conductivity dominance is inconsistent with the direction of silver line arrangement, and it is marked as a hidden anomalous grid.

[0085] Considering that, under ideal conditions, the current transport path in the coating is mainly dominated by the conductive network composed of silver nanowires, since the resistance to electron transport along the axial direction of the silver wires is much smaller than the resistance across the contact points between the silver wires, the current will preferentially transport along the direction of the silver wires themselves. For a grid cell, if the silver nanowires inside are most densely arranged and have the best connectivity in a certain direction, then that direction will naturally form the dominant channel for electron transport, thus exhibiting the lowest sheet resistance in that direction electrically.

[0086] Therefore, a theoretical correspondence can be established between microscopic morphology and macroscopic electrical properties: the dominant orientation angle of the silver nanowires within a grid cell should physically correspond precisely to their theoretically dominant conductivity direction, i.e., the direction of lowest resistance in actual measurements. In other words, the dominant orientation of the silver wires identified from the morphology image should be able to accurately predict the direction of low resistance in electrical measurements.

[0087] The core of anisotropy verification in this invention is to quantitatively test the above-mentioned theoretical correspondence. By calculating the deviation between the dominant orientation angle and the theoretical conductivity advantage angle reference determined by the low resistance direction, if the two directions deviate significantly, it indicates that the morphological structure has failed to be effectively transformed into the expected electrical performance. This phenomenon of the structure existing but not being electrically conductive is a latent anomaly.

[0088] Furthermore, S5, if the global trend verification fails, it enters the global anomaly screening stage, and identifies obvious abnormal grids by analyzing the regular deviations or independent anomalies between the line density and average sheet resistance of each grid unit.

[0089] When the global trend verification fails, it indicates that the coating as a whole does not exhibit the expected negative correlation between higher line density and lower average sheet resistance, suggesting a systematic process deviation that causes the coupling mechanism between morphology and electrical properties to fail. In this case, a global anomaly screening should be initiated to comprehensively identify any obvious abnormal meshes that deviate from the normal pattern.

[0090] Considering that anomalies may manifest in two forms, one is a deviation from the established pattern between morphology and electrical properties. Specifically, the purpose of global trend verification is overall qualitative analysis. It uses the rank correlation coefficient to determine whether the rule that higher linear density corresponds to lower sheet resistance still holds significantly across the entire coating scale. If the verification fails, it indicates that this rule has been broken at the overall level. Under this conclusion, the purpose of pattern deviation anomaly screening is local location. The core idea is that the disruption of the overall pattern must be caused by one or more local regions that violate this rule. To locate these regions, this invention shifts the analytical perspective from examining the overall correlation strength to evaluating the logical consistency between the morphology and electrical properties of each individual grid cell.

[0091] Another type is an independent anomaly where a certain indicator deviates extremely from its normal distribution, such as excessively low linear density or excessively high sheet resistance. Therefore, the comprehensive anomaly screening of this invention is carried out simultaneously from two dimensions to ensure full coverage detection of all types of explicit defects.

[0092] Referring to Figure 3, the process of full-domain anomaly screening corresponding to the deviation from the pattern is as follows: sort all grid cells of the full coating from high to low linear density and from low to high average sheet resistance.

[0093] For each grid cell, calculate its percentile rank in the online density sort and its percentile rank in the average sheet resistance sort.

[0094] The absolute value of the difference between the percentile rankings of two grid cells is defined as the consistency deviation of each grid cell. This deviation is used to characterize the degree to which the grid cell's own properties deviate from the basic law that high line density corresponds to low sheet resistance. It is compared with the allowable deviation. If the consistency deviation of a grid cell exceeds the allowable deviation, it is determined to be an obvious anomalous grid, that is, it is identified as a suspected defective cell that makes a significant contribution to the disruption of the overall conductivity law.

[0095] The allowable deviation mentioned above can be set based on the third quartile of the consistency deviation distribution of all grid cells in the full coating plus 1.5 times the interquartile range, reflecting the statistical upper limit of consistency deviation under normal process fluctuations.

[0096] In addition, the pattern deviation anomaly detection in the global anomaly screening is based on the idea of ​​ranking consistency analysis. When the global trend verification fails, it no longer relies on global correlation, but instead turns to local ranking comparison: for a normal grid, if its line density ranks high among all grids, then its average sheet resistance should rank low. If a grid has a high line density ranking, but its average sheet resistance ranking is also high, it means that the grid's own characteristics have serious internal contradictions and deviate greatly from basic physical expectations.

[0097] Therefore, by calculating the percentile ranking of each grid cell in the descending order of online density and the ascending order of average sheet resistance, and evaluating the consistency deviation between the two, the contribution of the grid to the disruption of the overall regularity can be quantified. The larger the deviation, the higher the suspicion level of the grid.

[0098] More specifically, the process of comprehensive anomaly screening corresponding to independent anomaly screening is as follows:

[0099] The linear density and average sheet resistance of all mesh cells in the full coating were plotted as frequency histograms.

[0100] Since histograms can clearly show the concentration trend of linear density or average sheet resistance throughout the coating, they facilitate the rapid identification of normal areas and abnormal edge areas, providing an intuitive threshold basis and judgment foundation for subsequent anomaly identification.

[0101] In the online density histogram, the linear density corresponding to the valley between the main distribution peak and the left tail is selected as the lower limit for judging linear density anomalies.

[0102] Understandably, there is usually a local minimum, or valley, between the main distribution peak and the low-value tail on its left in the online density histogram. The number of data points at this position is the smallest, marking the transition boundary between the normal structure region and the sparse abnormal region. Therefore, the corresponding line density value is set as the lower limit for anomaly judgment.

[0103] In the average sheet resistance histogram, the inflection point at the right tail of the main peak, where the high frequency transitions to the low frequency, is selected as the upper limit value for sheet resistance anomaly judgment.

[0104] Understandably, an inflection point or boundary from high frequency to low frequency can be identified at the beginning of the high value tail to the right of the main peak in the average sheet resistance histogram. This position represents the starting point of significant degradation in conductivity, so the sheet resistance value corresponding to it is used as the upper limit for anomaly judgment.

[0105] Perform dual-channel discrimination for each grid cell:

[0106] i) If the linear density is lower than the lower limit for judging linear density anomalies, it is marked as a mesh with obvious linear density anomalies;

[0107] ii) If the average sheet resistance is higher than the upper limit value for sheet resistance anomaly judgment, it is marked as a sheet resistance explicit anomaly grid.

[0108] Considering that line density reflects the density of silver lines per unit area, the higher the line density, the better. Too low a line density will lead to a serious lack of local conductivity. Therefore, anything below a certain threshold is considered abnormal, so it is set as the lower limit.

[0109] Another consideration is that the average sheet resistance characterizes the sheet resistance of the thin film. The lower the average sheet resistance, the stronger the conductivity. If it is too high, it indicates that the electrical performance is deteriorated. Therefore, anything above a certain threshold is considered abnormal, so it is set as the upper limit.

[0110] S6. Perform spatial clustering analysis on all abnormal meshes and aggregate them into defect regions with different attributes.

[0111] The specific details of the above steps are as follows: collect the spatial location information of all marked anomalous grids, including latent anomalous grids verified by anisotropy or explicit anomalous grids identified by global anomaly screening;

[0112] Spatial clustering is used to process the location information of abnormal meshes, and abnormal meshes that are spatially adjacent and have the same attribute are grouped into a defect region.

[0113] Given that a single anomalous grid may only reflect local measurement noise or minor fluctuations, while spatially continuous or clustered anomalous points are more likely to correspond to real process defects, this invention can restore discrete anomalies into coherent defect regions that conform to the actual physical scale through clustering and merging, which is closer to the real failure mode and also provides targeted location information for subsequent process optimization.

[0114] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0115] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0116] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0117] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0118] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting the conductivity of a conductive film nano-silver coating, characterized in that, include: A rectangular grid array was established on the surface of the nano-silver coating; Within each grid cell of the rectangular grid array, microscopic morphology images and transverse and longitudinal sheet resistance data of silver nanowires were collected. Based on the microscopic morphology images and transverse and longitudinal sheet resistance data of each grid cell, the line density, dominant orientation angle, average sheet resistance, and low resistance direction reflecting the electrical characteristics of the silver nanowires were extracted. Global trend verification was performed on all grid cells by analyzing the statistical correlation between line density and average sheet resistance. If the global trend verification passes, the anisotropy verification will proceed. By comparing the dominant orientation angle of the silver nanowires with the low resistance direction in each grid cell, hidden anomalous grids will be identified. If the global trend verification fails, the process proceeds to full-domain anomaly screening. By analyzing the regular deviations or independent anomalies between the line density and average sheet resistance of each grid cell, explicit anomalous grids are identified. Spatial clustering analysis is then performed on all anomalous grids to aggregate them into defect regions with different attributes.

2. The method for detecting the conductivity of a conductive film nano-silver coating as described in claim 1, characterized in that: The process of establishing a rectangular grid array on the surface of the nano-silver coating includes the following steps: establishing a two-dimensional rectangular coordinate system with the geometric center of the nano-silver coating as the origin, consisting of mutually orthogonal horizontal and vertical reference axes; dividing the surface along the horizontal and vertical reference axes at equal intervals to generate a set of non-overlapping, seamless grid units, forming a rectangular grid array covering the area to be tested of the coating; and assigning a spatial identifier to each grid unit, which contains its sequential position information on the horizontal and vertical reference axes.

3. The method for detecting the conductivity of a conductive film nano-silver coating as described in claim 1, characterized in that: The acquisition of microscopic morphology images and transverse and longitudinal sheet resistance data of silver nanowires is as follows: A composite detection device integrating a two-dimensional moving platform, optical imaging, and a four-probe sheet resistance measurement device is used; the two-dimensional moving platform drives the optical imaging device to move sequentially to the center area directly above each grid cell, and the surface image of the cell is acquired as the microscopic morphology image; at the same moving position, the device is switched to the four-probe sheet resistance measurement device, and the four-probe assembly is controlled to contact the center point of the grid cell. At the center point, measurement currents are applied sequentially along the predefined transverse and longitudinal measurement directions, and voltage signals are acquired. The transverse and longitudinal sheet resistance values ​​of the grid cell are then calculated.

4. The method for detecting the conductivity of a conductive film nano-silver coating as described in claim 1, characterized in that: The extraction of line density and dominant orientation angle, reflecting the morphological characteristics of silver nanowires, is performed as follows: The microscopic morphology image of each grid unit is segmented to separate the silver nanowire region from the background region; the segmented silver nanowire region is skeletonized to obtain a single-pixel-width silver nanowire skeleton network image; based on the silver nanowire skeleton network image, the sum of the actual physical lengths corresponding to all skeleton pixels within a unit area of ​​the grid unit is calculated and defined as the line density of the grid unit; statistical analysis is performed on the tangent directions of all skeleton pixels in the silver nanowire skeleton network image to construct a skeleton direction distribution histogram; the median value of the direction interval with the highest frequency distribution in the skeleton direction distribution histogram is determined as the dominant orientation angle of the silver nanowire arrangement within the grid unit.

5. The method for detecting the conductivity of a conductive film nano-silver coating as described in claim 1, characterized in that: The specific extraction process of the average sheet resistance and low resistance direction reflecting the electrical characteristics is as follows: the lateral sheet resistance value and the longitudinal sheet resistance value measured in each grid cell are arithmetically averaged to obtain the average sheet resistance reflecting the overall conductivity level of the cell; the lateral sheet resistance value and the longitudinal sheet resistance value in each grid cell are compared, and the direction with the smaller value is determined as the low resistance direction of the grid cell.

6. The method for detecting the conductivity of a conductive film nano-silver coating as described in claim 1, characterized in that: The global trend verification process is as follows: the line density of all grid cells is summarized to form the full coating line density sequence; the average sheet resistance of all grid cells is summarized to form the full coating average sheet resistance sequence; the rank correlation coefficient between the full coating line density parameter sequence and the full coating average sheet resistance sequence is calculated. The calculated rank correlation coefficient is compared with the strong negative correlation threshold; If the rank correlation coefficient is less than or equal to the strong negative correlation threshold, the overall conductive network of the coating is determined to conform to the basic physical law that the sheet resistance decreases as the silver line density increases, and the global trend verification is passed. Otherwise, the overall conductive network of the coating is determined to not conform to this basic physical law, and the global trend verification is not passed.

7. The method for detecting the conductivity of a conductive film nano-silver coating as described in claim 6, characterized in that: The anisotropy verification includes the following: determining the corresponding theoretical conductivity advantage angle reference based on the low resistance direction of each grid cell; Calculate the minimum angle difference between the dominant orientation angle of each grid cell and the theoretical conductivity advantage angle reference; compare the minimum angle difference with the configured angle tolerance threshold; if the minimum angle difference is greater than the angle tolerance threshold, it is determined that the grid cell has a structural imbalance where the conductivity advantage direction is inconsistent with the silver line arrangement direction, and it is marked as a hidden anomalous grid.

8. The method for detecting the conductivity of a conductive film nano-silver coating as described in claim 1, characterized in that: The regular deviation anomalies in the full-domain anomaly screening are as follows: sort all grid cells of the full coating from high to low linear density and from low to high average sheet resistance; for each grid cell, calculate its percentile ranking in the linear density sort and its percentile ranking in the average sheet resistance sort. The absolute value of the difference between the percentile rankings of the two is defined as the consistency deviation of each grid cell, and compared with the allowable deviation. If the consistency deviation of a grid cell exceeds the allowable deviation, it is determined to be an explicit anomalous grid.

9. The method for detecting the conductivity of a conductive film nano-silver coating as described in claim 1, characterized in that: The independent anomalies in the full-domain anomaly screening include the following: the line density and average sheet resistance of all grid cells in the full coating are plotted as frequency histograms; the line density corresponding to the valley position between the main distribution peak and the left tail in the line density histogram is selected as the lower limit value for line density anomaly judgment; the inflection point position of the high frequency to low frequency transition of the right tail of the main peak in the average sheet resistance histogram is selected as the upper limit value for sheet resistance anomaly judgment. For each grid cell, perform dual-channel discrimination: i) If the linear density is lower than the lower limit of the linear density anomaly judgment value, it is marked as a grid with obvious linear density anomaly; ii) If the average sheet resistance is higher than the upper limit of the sheet resistance anomaly judgment value, it is marked as a grid with obvious sheet resistance anomaly.

10. The method for detecting the conductivity of a conductive film nano-silver coating as described in claim 9, characterized in that: The spatial clustering analysis of all anomalous grids, aggregating them into defect regions with different attributes, includes the following: collecting the spatial location information of all marked anomalous grids, including latent anomalous grids verified by anisotropy or explicit anomalous grids identified by global anomaly screening. Spatial clustering is used to process the location information of abnormal meshes, and abnormal meshes that are spatially adjacent and have the same attribute are grouped into a defect region.

Images