RTF copper foil roughness prediction method based on structural parameters

Abstract

The invention discloses an RTF copper foil roughness prediction method based on structural parameters, and the method comprises the steps: 1, obtaining a copper nodule coverage rate x, a copper nodule size y and a copper nodule size variance z through the processing of an SEM image, and obtaining the copper foil roughness Rz through measurement; 2, determining the weights of the copper nodule coverage rate x, the copper nodule size y and the copper nodule size variance z; 3, fitting to obtain a uniform equation of the copper foil roughness Rz with the copper nodule coverage rate x, the copper nodule size y and the copper nodule size variance z; 4, fitting the three unary equations obtained in the step 3 through weights to obtain a nonlinear regression equation; and step 5, predicting the roughness according to the regression model of the copper foil roughness and the structure parameters of the copper foil to be measured. According to the method, the mathematical relationship between the copper foil structure and the roughness is explained, the copper foil roughness can be effectively predicted, and the estimation precision is high.

Description

Technical Field

[0001] The present invention relates to the field of copper foil testing, and in particular to a method for predicting the roughness of RTF copper foil based on structural parameters. Background Art

[0002] Electrolytic copper foil is one of the primary raw materials for printed circuit boards (PCBs). It serves as a carrier for signal and power transmission and communication in electronic products, significantly impacting PCB performance, manufacturing processes, costs, and service life. As electronic information technology enters a new era of rapid development, demand for high-frequency, high-speed PCBs is expected to experience a structural surge, placing even higher demands on the performance of copper foil.

[0003] Based on the application requirements of low roughness and high peel strength, relevant manufacturers have developed a reverse treated copper foil (RTF). This reverse treatment process of copper foil is relative to the traditional copper foil preparation process. For the prepared raw foil, the surface in contact with the cathode roller is usually called the smooth surface. The roughness of this surface is low, which can reduce the skin effect during high-frequency and high-speed signal transmission. However, its peel strength is low, making it difficult to prepare copper-clad laminates with hot pressing resins, and it is easy to cause falling off during the use of electronic circuit boards; and the surface of the raw foil in contact with the electrolyte is called the rough surface. The roughness of this surface is high. The traditional copper foil preparation process is to further roughen the rough surface of the copper foil and prepare a barrier layer and other post-processing. The reverse treatment process is to perform subsequent roughening treatment on the smooth surface of the copper foil, that is, to electro-deposit a layer of copper nodules with smaller grain size on the smooth surface of the copper foil again, which increases the specific surface area of the copper foil surface and improves 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. Most studies focus on simple structural descriptions, lacking detailed understanding of the interaction between the two. To address this issue, the present invention constructs a regression equation between roughness and structural parameters, mathematically explaining the influence of structure on roughness and enabling prediction of copper foil roughness based on structure. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for predicting the roughness of RTF copper foil based on structural parameters to solve the shortcomings of the above-mentioned prior art.

[0006] To achieve the above object, 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: The copper nodule coverage x, copper nodule size y, and copper nodule size variance z are obtained by processing the SEM image, and the copper foil roughness Rz is obtained 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, fitting to obtain a single-variable equation of the copper foil roughness Rz and the copper nodule coverage x, copper nodule size y, and copper nodule size variance z;

[0010] Step 4: Fit the three univariate equations obtained in step 3 by weight to obtain a nonlinear regression equation;

[0011] Step 5: Predict the roughness based on the regression model of the copper foil roughness and the structural parameters of the copper foil to be tested.

[0012] Furthermore, in step 1, ImageJ software is used to process the SEM image of the RTF copper foil and calculate the structural parameters, including the copper nodule coverage x, the copper nodule size y, and the copper nodule size variance z; during the calculation, different thresholds are adjusted for the SEM image as required; and at the same time, a roughness tester is used to measure the roughness Rz of the RTF copper foil to obtain a corresponding set of roughness values.

[0013] Furthermore, the threshold value is 25-35 when calculating the copper nodule coverage, and the threshold value is 60-80 when calculating the copper nodule size and variance.

[0014] Furthermore, in step 2, the weights of the copper nodule coverage x, the copper nodule size y, and the 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 take the single structural parameter obtained in step 1 as a variable and the roughness obtained in step 1 as a dependent variable to obtain multiple models, and R is selected. 2 The largest model is used as the calculation 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 include: linear function, logarithmic function, inverse function, quadratic function, cubic function, composite function, power function, S curve, growth curve, exponential function, and logistic function.

[0020] Furthermore, in step 4, the cross term axy+bxz+cyz and the constant term d of the undetermined coefficients are introduced, and the nonlinear multiple regression equation is constructed using the nonlinear regression function of the SPSS software.

[0021] Furthermore, the constructed nonlinear multiple regression equation is:

[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] Among them, a, b, c, and d are the introduced unknown coefficients. By substituting the roughness Rz, copper nodule coverage x, copper nodule size y, and copper nodule size variance z of multiple groups of different RTF copper foils into the above formula, the values of these four unknown 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, the fitting degree of the final multiple regression equation was analyzed to obtain the R 2 value to verify the fitting effect.

[0026] The beneficial effects of the present invention are:

[0027] (1) The present invention processes the scanning electron microscope (SEM) images of the RTF copper foil surface to extract key structural parameters such as copper nodule coverage, copper nodule size and size variance, and establishes a mathematical model between them and roughness, filling the gap in the existing technology that roughness and microstructure cannot be directly quantified, and providing a scientific basis for the structural regulation of copper foil.

[0028] (2) Through SPSS curve estimation and nonlinear regression analysis, a multivariate regression equation with high fitting degree (R2=0.935) between roughness and multiple structural parameters was constructed, which can efficiently predict the roughness of copper foil under different structural combinations and has good versatility and engineering guidance value.

[0029] (3) The analysis method and regression model provided by the present invention can be used as tools for copper foil roughening process optimization, performance prediction and quality control. They have good industrial adaptability and promotion value, help promote the manufacture of high-performance electronic copper foil materials, and have engineering promotion prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of the RTF copper foil roughness prediction method based on structural parameters of the present invention.

[0031] Figure 2 This is the SEM image of the RTF copper foil prepared at a current density of 700 A / m2 in Example 1.

[0032] Figure 3 This is the coverage diagram of the RTF copper foil prepared at a current density of 700 A / m2 in Example 1.

[0033] Figure 4 This is a diagram of the copper nodule size of the RTF copper foil prepared at a current density of 700 A / m2 in Example 1.

[0034] Figure 5 This is the error diagram between the roughness experimental value and the predicted value in Example 1.

[0035] Figure 6 These are SEM images of the RTF copper foil prepared at current densities of 1300, 500, and 1300 A / m2 in Example 2.

[0036] Figure 7 This is the coverage diagram of the RTF copper foil prepared at current densities of 1300, 500, and 1300 A / m2 in Example 2.

[0037] Figure 8 This is a diagram of the copper nodule size of the RTF copper foil prepared at current densities of 1300, 500, and 1300 A / m2 in Example 2.

[0038] Figure 9 This is the error diagram between the roughness experimental value and the predicted value in Example 2. DETAILED DESCRIPTION

[0039] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0040] like Figure 1 , a RTF copper foil roughness prediction method based on structural parameters, comprising 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; adjust the threshold of the SEM image according to the needs during the calculation. The threshold for calculating coverage is generally 25-35, and the threshold for calculating copper nodule size and variance is generally around 60-80; use a roughness tester to measure the roughness Rz of the RTF copper foil and obtain a corresponding set of roughness values;

[0042] Step 2: Using the analytic hierarchy process (AHP) and expert evaluation, we determined that for roughness, coverage x and nodule size y are significantly more important, slightly more important than size variance z, and size variance z is slightly more important than nodule size y. Therefore, after multiple fittings, we determined the weights for coverage x, nodule size y, and nodule size variance z to be 0.637, 0.105, and 0.258, respectively.

[0043] Step 3: Use the curve estimation function of SPSS software, take the single structural parameter obtained in step 1 as the variable, take the roughness obtained in step 1 as the dependent variable, obtain multiple models, and select the model with the largest R2 as the calculation equation:

[0044] The equation of roughness Rz-copper nodule coverage x 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 equation of roughness Rz-copper nodule size variance z 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, taking into account the mutual influence of structural parameters, the cross term axy+bxz+cyz of the undetermined coefficients and the constant term d are introduced, and the nonlinear regression function of SPSS software is 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] Among them, a, b, c, and d are the introduced unknown coefficients. By substituting the relevant parameters into the four unknown coefficients, the values of these four unknown 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 fitting analysis on the multiple regression equation obtained in step 4 to obtain the R 2 The fitting effect is good.

[0052] In step 3, the multiple models include: linear, logarithmic, inverse, quadratic, cubic, composite, power, S, growth, exponential, and logistic functions.

[0053] Table 1 Model function types and their expressions

[0054]

[0055] The above method is further explained with specific cases:

[0056] Example 1

[0057] Example 1 of the present invention provides a structural parameter-based RTF copper foil roughness prediction method, applicable to copper foil roughness testing scenarios. Nine different copper foils (300, 400, 500, 600, 700, 800, 900, 1000, and 1100 A / m²) were selected for testing. SEM images and roughness values were obtained, and structural parameters were calculated. Substituting these structural parameters into the aforementioned regression equation yielded the corresponding roughness predictions.

[0058] Specifically, taking the sample prepared at a current density of 720A / m2 as an example, the average roughness measured by the roughness meter is 2.893μm. Figure 2 This is the SEM image of the TF copper foil surface.

[0059] Will Figure 2 After binarization in Imagej software, the threshold was adjusted to 30, and the Figure 3 Coverage map, the calculated coverage is 93.54%.

[0060] Will Figure 2 After binarization in Imagej software, the threshold was adjusted to 70, and the Figure 4 Copper nodule size diagram, the average size of the copper nodule is calculated to be 1.18μm, and the copper nodule size variance is 0.302.

[0061] Please refer to Figure 5To verify the error between the model predictions and experimental values, the experimental data were compared with the multivariate nonlinear regression model. It can be seen that the regression equation predictions follow 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%, demonstrating a high degree of agreement.

[0062] Example 2

[0063] RTF copper foil was prepared by a three-step method. The current densities of the three steps were selected from "500, 900, 1300A / m2", "300, 500, 700A / m2", and "1300, 2100, 2900A / m2", respectively. Twenty-seven groups of different copper foils were prepared by roughening at different current densities. SEM images and roughness were obtained by testing, and the structural parameters were calculated. After substituting the structural parameters into the above regression equation, the corresponding roughness prediction values were obtained.

[0064] Specifically, the sample prepared in three steps with current density of 1260, 500, and 1260 A / m2 was selected as an example, and the average roughness was measured by a roughness meter to be 3.761 μm. Figure 6 This is the SEM image of the RTF copper foil surface.

[0065] Will Figure 6 After binarization in Imagej software, the threshold was adjusted to 30, and the Figure 7 Coverage map, the calculated coverage is 89.98%.

[0066] Will Figure 6 After binarization in Imagej software, the threshold was adjusted to 70, and the Figure 8 Copper nodule size diagram, the average size of the copper nodule is calculated to be 1.32μm, and the variance of the copper nodule size is 0.606.

[0067] Please refer to Figure 9 ,In order to verify the error between the model prediction value and the experimental value, the experimental data are compared with the multivariate nonlinear regression model, such as Figure 2 As shown in Figure 2, it can be seen that the predicted values of the regression equation are roughly consistent with the experimental data trends, with a maximum error of 15%, a minimum error of 1.76%, and an average relative error of 7.86%, which shows a certain degree of consistency.

[0068] The above shows and describes 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 form, and all technical solutions obtained by equivalent replacement and other methods fall within the scope of protection of the present invention.

[0069] The parts not involved in the present invention are the same as the existing technology or can be implemented by using the existing technology.

Claims

1. A method for predicting the roughness of RTF copper foil based on structural parameters, characterized in that: The steps include: Step 1: The copper nodule coverage x, copper nodule size y, and copper nodule size variance z are obtained by processing the SEM image, and the copper foil roughness Rz is obtained by measurement; Step 2, determine the weights of copper nodule coverage x, copper nodule size y, and copper nodule size variance z; Step 3, fitting to obtain a single-variable equation of the copper foil roughness Rz and the copper nodule coverage x, copper nodule size y, and copper nodule size variance z; Step 4: Fit the three univariate equations obtained in step 3 by weight to obtain a nonlinear regression equation; Step 5: Predict the roughness based on the regression model of the copper foil roughness and the structural parameters of the copper foil to be tested.

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 and structural parameters are calculated, including copper nodule coverage x, copper nodule size y, and copper nodule size variance z; different thresholds are adjusted for the SEM image as required during the calculation; and the roughness Rz of the RTF copper foil is measured using a roughness tester to obtain a corresponding 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 value for calculating the copper nodule coverage is 25-35, and the threshold value for calculating the 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 the copper nodule coverage x, the copper nodule size y, and the 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: In step 3, the curve estimation function of SPSS software is used to take the single structural parameter obtained in step 1 as a variable and the roughness obtained in step 1 as a dependent variable to obtain multiple models. 2 The largest model is used as the calculation equation.

6. The RTF copper foil roughness prediction method based on structural parameters according to claim 5, characterized in that: The equation of roughness Rz-copper nodule coverage x is Rz=-14.133x3+25.714x-9.412; The roughness Rz-copper nodule size y equation is Rz = 20.424y3-87.457y+72.659; The roughness Rz-copper nodule size variance z equation is Rz = -2.034z3 + 3.837z + 1.

831.

7. The RTF copper foil roughness prediction method based on structural parameters according to claim 5, characterized in that: Multiple models include: linear function, logarithmic function, inverse function, quadratic function, cubic function, composite function, power function, S curve, growth curve, exponential function, and logistic function.

8. The RTF copper foil roughness prediction method based on structural parameters according to claim 1, characterized in that: In step 4, the cross term axy+bxz+cyz and the constant term d of the undetermined coefficients are introduced, and the nonlinear multiple regression equation is constructed using the nonlinear regression function of the SPSS software.

9. The method for predicting RTF copper foil roughness based on structural parameters according to claim 8, characterized in that: The constructed nonlinear multiple regression equation is: 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; Among them, a, b, c, and d are the introduced unknown coefficients. By substituting the roughness Rz, copper nodule coverage x, copper nodule size y, and copper nodule size variance z of multiple groups of different RTF copper foils into the above formula, the values of these four unknown coefficients are calculated, and finally the final multiple regression equation is obtained: 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.

10. The method for predicting RTF copper foil roughness based on structural parameters according to claim 9, characterized in that: The fitting degree of the final multiple regression equation was analyzed to obtain the R 2 value to verify the fitting effect.

Images