RTF Copper Foil Roughness Prediction Method Based on Structural Parameters
By constructing a nonlinear multiple regression equation for the structural parameters of copper foil, the research on the relationship between the roughness and performance of RTF copper foil is insufficient, and efficient roughness prediction is achieved, which has good prospects for industrial application.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JIANGDONG ELECTRONIC MATERIALS CO LTD
- Filing Date
- 2025-05-09
- Publication Date
- 2026-08-04
AI Technical Summary
In the existing technology, there is little research on the relationship between the roughness and performance of RTF copper foil, and there is a lack of research on the specific interaction relationship, making it difficult to effectively predict its roughness through structural parameters.
By constructing weights for copper nodule coverage, copper nodule size, and size variance, and using ImageJ software to process SEM images and combining them with SPSS software, a nonlinear multiple regression equation was established to predict the roughness of copper foil.
It achieves high-fit roughness prediction, fills the gap in the quantitative correlation between roughness and microstructure, provides a scientific basis for the structural control of copper foil, and has good industrial adaptability and promotion value.
Smart Images
Figure CN120507385B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of copper foil testing, and more specifically to an RTF method for predicting the roughness of copper foil based on structural parameters. Background Technology
[0002] Electrolytic copper foil is one of the main raw materials for printed circuit boards (PCBs). It serves as the carrier for signal and power transmission and communication in electronic products, significantly impacting PCB performance, manufacturing processes, manufacturing costs, and lifespan. With the rapid development of electronic information technology, the demand for high-frequency, high-speed PCBs will experience structural and rapid growth, placing higher demands on the performance of copper foil.
[0003] Driven by the application requirements of low roughness and high peel strength, relevant manufacturers have developed a reverse-treated copper foil (RTF). This reverse treatment process is relative to the traditional copper foil manufacturing process. For the prepared green foil, the surface in contact with the cathode roller is usually called the smooth surface. This surface has low roughness, which can reduce the skin effect during high-frequency, high-speed signal transmission, but its peel strength is low, making it difficult to hot-press resin to prepare copper-clad laminates, and it is prone to peeling during use in electronic circuit boards. The surface in contact with the electrolyte is called the rough surface, which has higher roughness. Traditional copper foil manufacturing processes further roughen the rough surface of the copper foil, preparing a barrier layer and other post-processing steps. The reverse treatment process, however, performs a subsequent roughening treatment on the smooth surface of the copper foil, that is, electrodepositing a copper nodule layer with smaller grain size on the smooth surface of the copper foil, increasing the specific surface area of the copper foil and improving the peel strength.
[0004] The surface structure of RTF copper foil determines its roughness, but existing research on the relationship between RTF copper foil roughness and performance is limited, mostly focusing on simple descriptions of the structure without exploring the specific interaction between the two. To address this issue, this invention constructs a regression equation between roughness and structural parameters, mathematically explaining the influence of structure on roughness and enabling the prediction of copper foil roughness based on its structure. Summary of the Invention
[0005] The purpose of this invention is to provide an RTF copper foil roughness prediction method based on structural parameters to overcome the shortcomings of the existing technology.
[0006] To achieve the above objectives, the technical solution adopted by the present invention is: a method for predicting the roughness of RTF copper foil based on structural parameters, which includes the following steps:
[0007] Step 1: Obtain the copper nodule coverage x, copper nodule size y, and copper nodule size variance z by processing the SEM image, and obtain the copper foil roughness Rz by measurement;
[0008] Step 2: Determine the weights of copper nodule coverage x, copper nodule size y, and copper nodule size variance z;
[0009] Step 3: Fit the univariate equations of copper foil roughness Rz with copper nodule coverage x, copper nodule size y, and copper nodule size variance z respectively;
[0010] Step 4: Fit the three univariate equations obtained in Step 3 to obtain a nonlinear regression equation by weighting them.
[0011] Step 5: Based on the regression model of copper foil roughness and the structural parameters of the copper foil to be tested, predict the roughness.
[0012] Furthermore, in step 1, the SEM image of the RTF copper foil is processed using ImageJ software to calculate structural parameters, including copper nodule coverage x, copper nodule size y, and copper nodule size variance z. During the calculation, different thresholds are adjusted for the SEM image as needed. At the same time, the roughness Rz of the RTF copper foil is measured using a roughness tester to obtain a set of corresponding roughness values.
[0013] Furthermore, the threshold for calculating copper nodule coverage is 25-35, and the threshold for calculating copper nodule size and variance is 60-80.
[0014] Furthermore, in step 2, the weights of copper nodule coverage x, copper nodule size y, and copper nodule size variance z are 0.637, 0.105, and 0.258, respectively.
[0015] Furthermore, in step 3, the curve estimation function of SPSS software is used to obtain multiple models by taking the single structural parameter obtained in step 1 as a variable and the roughness obtained in step 1 as a dependent variable, and then selecting R... 2 The largest model is used as the computational equation.
[0016] Furthermore, the roughness Rz-copper nodule coverage x equation is Rz=-14.133x3+25.714x-9.412;
[0017] The roughness Rz-copper nodule size y equation is Rz=20.424y3-87.457y+72.659;
[0018] The roughness Rz-copper nodule size variance z equation is Rz=-2.034z3+3.837z+1.831.
[0019] Furthermore, multiple models are included, such as: linear functions, logarithmic functions, inverse functions, quadratic functions, cubic functions, composite functions, power functions, S-curves, growth curves, exponential functions, and logistic functions.
[0020] Furthermore, in step 4, the cross term axy+bxz+cyz with undetermined coefficients and the constant term d are introduced, and the nonlinear regression function of SPSS software is used to construct a nonlinear multiple regression equation.
[0021] Furthermore, the constructed nonlinear multiple regression equation is as follows:
[0022] Rz=0.637*(-14.133x3+25.714x-9.412)+0.105*(20.424y3-87.457y+72.659)+0.258*(-2.034z3+3.837z+1.831)+axy+bxz+cyz+d;
[0023] Where a, b, c, and d are the undetermined coefficients introduced. By substituting the roughness Rz, copper nodule coverage x, copper nodule size y, and copper nodule size variance z of multiple sets of different RTF copper foils into the above formula, the values of these four undetermined coefficients are calculated, and finally the final multiple regression equation is obtained:
[0024] Rz=0.637*(-14.133x3+25.714x-9.412)+0.105*(20.424y3-87.457y+72.6 59)+0.258*(-2.034z3+3.837z+1.831)-0.085xy-1.683xz+1.380yz+0.072.
[0025] Furthermore, a goodness-of-fit analysis is performed on the final multiple regression equation to obtain the R-squared value of the regression equation. 2 The value was used to verify the fitting effect.
[0026] The beneficial effects of this invention are:
[0027] (1) This invention extracts key structural parameters such as copper nodule coverage, copper nodule size and size variance by processing scanning electron microscope (SEM) images of RTF copper foil surface, and establishes a mathematical model between them and roughness, filling the gap in the prior art where roughness and microstructure cannot be directly quantified and correlated, and providing a scientific basis for copper foil structure control.
[0028] (2) By estimating the SPSS curve and performing nonlinear regression analysis, a multivariate regression equation with a high degree of fit (R2 = 0.935) between roughness and multiple structural parameters was constructed. This equation can efficiently predict the roughness of copper foil under different structural combinations and has good versatility and engineering guidance value.
[0029] (3) The analytical method and regression model provided by this invention can be used as tools for optimizing copper foil roughening process, predicting performance and controlling quality. They have good industrial adaptability and promotion value, which helps to promote the manufacturing of high-performance electronic copper foil materials and has engineering promotion prospects. Attached Figure Description
[0030] Figure 1 This is a flowchart of the RTF copper foil roughness prediction method based on structural parameters according to the present invention.
[0031] Figure 2 The image shows the SEM image of the RTF copper foil prepared with a current density of 700 A / m² in Example 1.
[0032] Figure 3 The image shows the coverage of the RTF copper foil prepared with a current density of 700 A / m² in Example 1.
[0033] Figure 4 This is a diagram showing the copper nodule dimensions of the RTF copper foil prepared with a current density of 700 A / m² in Example 1.
[0034] Figure 5 This is a graph showing the error between the experimental and predicted roughness values in Example 1.
[0035] Figure 6 The image shows SEM images of the RTF copper foils prepared with current densities of 1300, 500, and 1300 A / m² in Example 2.
[0036] Figure 7 The diagram shows the coverage of RTF copper foils prepared with current densities of 1300, 500, and 1300 A / m2 in Example 2.
[0037] Figure 8 The diagram shows the copper nodule dimensions of the RTF copper foils prepared with current densities of 1300, 500, and 1300 A / m2 in Example 2.
[0038] Figure 9 This is a graph showing the error between the experimental and predicted roughness values in Example 2. Detailed Implementation
[0039] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.
[0040] like Figure 1 A method for predicting the roughness of RTF copper foil based on structural parameters includes the following steps:
[0041] Step 1: Use ImageJ software to process the SEM image of the RTF copper foil and calculate the structural parameters, including copper nodule coverage x, copper nodule size y, and copper nodule size variance z. During the calculation, adjust the threshold of the SEM image according to the requirements. When calculating the coverage, the threshold is generally 25-35, and when calculating the copper nodule size and variance, it is generally around 60-80. Use a roughness tester to measure the roughness Rz of the RTF copper foil and obtain a set of corresponding roughness values.
[0042] Step 2: Using the analytic hierarchy process (AHP) and combined with expert evaluations, for roughness, coverage (x) and copper nodule size (y) are significantly more important, and slightly more important than size variance (z). Size variance (z) is slightly more important than copper nodule size (y). Therefore, after multiple fitting iterations, the weights of coverage (x), copper nodule size (y), and copper nodule size variance (z) are determined to be 0.637, 0.105, and 0.258, respectively.
[0043] Step 3: Using the curve estimation function of SPSS software, take the single structural parameter obtained in Step 1 as the variable and the roughness obtained in Step 1 as the dependent variable to obtain multiple models. Select the model with the largest R² as the calculation equation:
[0044] The roughness Rz-copper nodule coverage x equation is Rz=-14.133x3+25.714x-9.412;
[0045] The roughness Rz-copper nodule size y equation is Rz=20.424y3-87.457y+72.659;
[0046] The roughness Rz-copper nodule size variance z equation is Rz=-2.034z3+3.837z+1.831;
[0047] Step 4: Based on the weights obtained in Step 2 and the multiple regression equations obtained in Step 3, considering the influence between the structural parameters, the interaction term axy+bxz+cyz with undetermined coefficients and the constant term d are introduced. The nonlinear regression function of SPSS software is then used to construct a multiple regression equation:
[0048] Rz=0.637*(-14.133x3+25.714x-9.412)+0.105*(20.424y3-87.457y+72.659)+0.258*(-2.034z3+3.837z+1.831)+axy+bxz+cyz+d;
[0049] Where a, b, c, and d are the introduced undetermined coefficients. By substituting the relevant parameters, the values of these four undetermined coefficients are calculated, and finally, the final multiple regression equation is obtained:
[0050] Rz=0.637*(-14.133x3+25.714x-9.412)+0.105*(20.424y3-87.457y+72.6 59)+0.258*(-2.034z3+3.837z+1.831)-0.085xy-1.683xz+1.380yz+0.072;
[0051] Step 5: Perform a goodness-of-fit analysis on the multiple regression equation obtained in Step 4 to obtain the R-squared value of the regression equation. 2 The value is 0.935, indicating a good fit.
[0052] In step 3, multiple models include: linear, logarithmic, inverse, quadratic, cubic, composite, power, S-shaped, growth, exponential, and logistic functions.
[0053] Table 1 Model function types and their expressions
[0054]
[0055] The above methods will be further explained using specific examples:
[0056] Example 1
[0057] Embodiment 1 of the present invention is an RTF copper foil roughness prediction method based on structural parameters, which can be applied to copper foil roughness testing scenarios. Nine different copper foils with roughness values of 300, 400, 500, 600, 700, 800, 900, 1000, and 1100 A / m² were selected, and SEM images and roughness were obtained. The structural parameters were calculated, and the corresponding roughness prediction values were obtained by substituting the structural parameters into the above regression equation.
[0058] Specifically, taking a sample prepared with a current density of 720 A / m² as an example, the average roughness measured using a roughness tester was 2.893 μm. Figure 2 This is a SEM image of the TF copper foil surface.
[0059] Will Figure 2 After binarization in Imagej software, the threshold is adjusted to 30, resulting in... Figure 3 The coverage map shows a calculated coverage rate of 93.54%.
[0060] Will Figure 2 After binarization in Imagej software, the threshold is adjusted to 70, resulting in... Figure 4 The copper nodule size diagram shows that the average size of the copper nodule is 1.18 μm and the variance of the copper nodule size is 0.302.
[0061] Please refer to Figure 5To verify the error between the model's predicted values and the experimental values, the experimental data were compared with the multivariate nonlinear regression model. It can be seen that the predicted values of the regression equation show the same trend as the experimental data, with a maximum error of 4.60%, a minimum error of 1.12%, and an average relative error of 2.32%, indicating a high degree of agreement.
[0062] Example 2
[0063] RTF copper foil was prepared using a three-step method. The current densities for the three steps were selected from "500, 900, 1300 A / m2", "300, 500, 700 A / m2", and "1300, 2100, 2900 A / m2", respectively. Twenty-seven different groups of copper foil were prepared by roughening with different current densities. SEM images and roughness were obtained by testing, and structural parameters were calculated. After substituting the structural parameters into the above regression equation, the corresponding roughness prediction values were obtained.
[0064] Specifically, taking samples prepared with current densities of 1260, 500, and 1260 A / m² in three steps as an example, the average roughness measured using a roughness meter was 3.761 μm. Figure 6 This is a SEM image of the RTF copper foil surface.
[0065] Will Figure 6 After binarization in Imagej software, the threshold is adjusted to 30, resulting in... Figure 7 The coverage map shows a calculated coverage rate of 89.98%.
[0066] Will Figure 6 After binarization in Imagej software, the threshold is adjusted to 70, resulting in... Figure 8 The copper nodule size diagram shows that the average size of the copper nodule is 1.32 μm and the variance of the copper nodule size is 0.606.
[0067] Please refer to Figure 9 To verify the error between the model's predicted values and experimental values, the experimental data were compared with the multivariate nonlinear regression model, such as... Figure 2 As shown, the predicted values of the regression equation have roughly the same trend as the experimental data, with a maximum error of 15%, a minimum error of 1.76%, and an average relative error of 7.86%, indicating a certain degree of agreement.
[0068] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the scope of protection of the present invention in any way, and all technical solutions obtained by equivalent substitution or other means fall within the scope of protection of the present invention.
[0069] All parts not covered in this invention are the same as or can be implemented using existing technologies.
Claims
1. A method for predicting the roughness of RTF copper foil based on structural parameters, characterized in that, Includes the following steps: Step 1: Obtain the copper nodule coverage x, copper nodule size y, and copper nodule size variance z by processing the SEM image, and obtain the copper foil roughness Rz by measurement; Step 2: Determine the weights of copper nodule coverage x, copper nodule size y, and copper nodule size variance z; Step 3: Fit the univariate equations of copper foil roughness Rz with copper nodule coverage x, copper nodule size y, and copper nodule size variance z. Specifically, use the curve estimation function of SPSS software, taking the individual structural parameter obtained in Step 1 as the variable and the roughness obtained in Step 1 as the dependent variable, to obtain multiple models. Select Rz as the dependent variable. 2 The largest model serves as the computational equation, which includes multiple models such as: linear functions, logarithmic functions, inverse functions, quadratic functions, cubic functions, composite functions, power functions, S-curves, growth curves, exponential functions, and logistic functions; Step 4: Fit the three univariate equations obtained in Step 3 to obtain a nonlinear regression equation by weighting them; specifically, introduce the cross term axy+bxz+cyz with undetermined coefficients and the constant term d, and use the nonlinear regression function of SPSS software to construct a nonlinear multiple regression equation. Step 5: Based on the regression model of copper foil roughness and the structural parameters of the copper foil to be tested, predict the roughness.
2. The RTF copper foil roughness prediction method based on structural parameters according to claim 1, characterized in that, In step 1, the SEM image of the RTF copper foil is processed using ImageJ software to calculate structural parameters, including copper nodule coverage x, copper nodule size y, and copper nodule size variance z. During the calculation, different thresholds are adjusted for the SEM image as needed. At the same time, the roughness Rz of the RTF copper foil is measured using a roughness tester to obtain a set of roughness values.
3. The RTF copper foil roughness prediction method based on structural parameters according to claim 2, characterized in that, The threshold for calculating copper nodule coverage is 25-35, and the threshold for calculating copper nodule size and variance is 60-80.
4. The RTF copper foil roughness prediction method based on structural parameters according to claim 1, characterized in that, In step 2, the weights of copper nodule coverage x, copper nodule size y, and copper nodule size variance z are 0.637, 0.105, and 0.258, respectively.
5. The RTF copper foil roughness prediction method based on structural parameters according to claim 1, characterized in that, A goodness-of-fit analysis was performed on the final multiple regression equation to obtain the R-squared value of the regression equation. 2 The value was used to verify the fitting effect.