Method for predicting geometric and performance characteristics of magnetron sputtering aluminum-tin coating

Abstract

The invention belongs to the technical field of physical vapor deposition forming precision, and discloses a method for predicting geometric and performance characteristics of a magnetron sputtering aluminum-tin coating, which comprises the following steps of: investigating the influence of technological parameters such as sputtering power, substrate bias voltage, sputtering air pressure and target-substrate distance on the performance of an aluminum-tin alloy coating on the basis of an independent variable control strategy system; the shape change rule of the aluminum-tin coating under the action of different process parameters is disclosed; based on a response curved surface experiment design theory and a multi-factor optimization theory, a coating process-performance mapping mechanism under multiple process parameters is explored; establishing a characteristic database containing process parameters and coating geometry and performance characteristics; and optimizing coating process parameters based on the coating characteristic target value, clarifying a coating shape quality change rule under the process coupling effect, and optimizing the process parameters. According to the method, the mapping model of the process parameters and the geometric characteristics of the coating can be quickly and accurately established, the process cost is low, the efficiency is high, and the precision is controllable.

Description

Technical Field

[0001] The invention belongs to the technical field of physical vapor deposition forming accuracy, and relates to a method for quickly predicting geometry and performance characteristics of a magnetron sputtering coating. Background Art

[0002] Magnetron sputtering technology is a process in which voltage is applied between electrodes under vacuum conditions to generate glow discharge, which generates electrons. During the process of movement, the electrons collide with the gas introduced, ionizing the gas molecules and generating a large number of gas ions and new electrons. The gas ions are accelerated to bombard the target material under the action of the electric field, causing the target atoms to sputter out, and the target atoms are finally deposited on the substrate to form a coating. The coating prepared by magnetron sputtering has outstanding characteristics such as good adhesion, uniform thickness, high-purity thin film and low-temperature deposition. In addition, magnetron sputtering has a fast growth rate, is easy to automate, can be used for large-scale production, and has broad application prospects.

[0003] Theoretically, magnetron sputtering technology can form coatings of different materials, but there are still many problems to be solved in actual production, which limits the application scope of magnetron sputtering deposition technology. At present, there are three main problems in magnetron sputtering technology. The first is the equipment: magnetron sputtering equipment is usually more complex, with high equipment cost and difficulty in maintenance; the second is the material aspect. For aluminum-tin coatings, tin has a low melting point and is easy to volatilize, which leads to tin loss or non-uniform distribution at high temperatures. Controlling the tin content and distribution in deposition is a major problem, which usually needs to be adjusted by optimizing the deposition parameters; the third is the process aspect: process parameters (sputtering power, substrate bias, sputtering gas pressure, target-substrate distance and other process parameters) directly affect the morphological quality of the deposited layer (geometric dimensions, microhardness, coating adhesion, etc.). With the current equipment, the problem of quickly establishing an accurate coating morphological quality prediction model is affecting the development and application of magnetron sputtering deposition technology and has become a key issue in improving the quality of coating formation.

[0004] In the field of magnetron sputtering deposition of aluminum-tin alloy coatings, there are currently studies related to parameters such as substrate temperature, deposition gas pressure, substrate bias and sputtering power. During the magnetron sputtering process, if the target-substrate distance is reduced, the distance that the atoms sputtered from the target move to the substrate becomes shorter, which reduces the probability of collision with the working gas, resulting in an increase in the number of atoms moving to the substrate and an increase in the coating thickness. However, there is currently no research on the effect of the target-substrate distance parameter on the key morphological quality of the coating. At the same time, the current relationship between the process parameters and morphological quality of magnetron sputtering aluminum-tin alloy coatings is mostly based on single factor analysis, lacking research on the interaction relationship of multiple factors and exploration of the mapping relationship of morphological quality under complex multi-process coupling, and it is impossible to optimize the process for complex deposition processes. The interaction between multiple processes will bring great changes to the key morphological quality of the coating. The process based on multi-experimental data exploration is expensive and inefficient, and the prediction accuracy is uncontrollable. At present, there is no effective method to quickly and accurately establish a prediction model for coating geometry and performance characteristics. Therefore, it is necessary to propose a prediction method for the geometry and performance characteristics of aluminum-tin alloy coatings deposited by magnetron sputtering under the action of multi-process coupling, clarify the law of change of coating morphological quality under process coupling and optimize the process parameters to complete the rapid and accurate prediction of coating geometric characteristics, thereby obtaining high-performance aluminum-tin alloy coatings. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention expands the variation law of the geometry and performance characteristics of the aluminum-tin alloy coating deposited by magnetron sputtering under the process change of the target-substrate distance, provides a method for predicting the geometry and performance characteristics of the aluminum-tin alloy coating deposited by magnetron sputtering, establishes a prediction model of process parameters and the geometry and performance characteristics of the aluminum-tin coating, and realizes the control of the model accuracy. At the same time, the prediction model has strong adaptability and is not restricted by the size and properties of the substrate or feed material. This solves the problems of high process cost, low efficiency and uncontrollable prediction accuracy of the original method for predicting the geometry and performance characteristics of the aluminum-tin alloy coating deposited by magnetron sputtering based on multiple experiments, and lays the foundation for the subsequent establishment of process planning and intelligent manufacturing that can realize rapid automation in the deposition manufacturing process.

[0006] The technical solution of the present invention:

[0007] A method for predicting the geometry and performance characteristics of magnetron sputtering aluminum-tin alloy coatings. In the process of magnetron sputtering aluminum-tin alloy coatings, the target-substrate distance (the distance between the target and the substrate) is a key process parameter that has not been fully studied. The target-substrate distance may have an important impact on the deposition quality, deposition rate and performance of the coating, but there are still few systematic studies on this parameter. This method aims to use the target-substrate distance as the research entry point to systematically investigate the effects of four process parameters, namely sputtering power, substrate bias, sputtering gas pressure and target-substrate distance, on the performance of aluminum-tin alloy coatings. The response surface method is used to explore the interaction of these parameters and their mapping relationship with the coating performance, and on this basis, the process is optimized to improve the performance of the coating. This method is based on the structural mapping relationship between parameters such as target-substrate distance and coating performance; based on the response surface experimental design theory and multi-factor optimization theory, a feature database containing process parameters and the geometry and performance characteristics of the deposited layer is established. Among them, the process parameters include sputtering power, sputtering gas pressure, substrate bias and target-substrate distance, the coating geometric characteristics are layer thickness and density, and the performance characteristics mainly include hardness and bonding strength; on the basis of sputtering experiments based on response surface design, the significant relationship between each process parameter and the coating layer thickness, density and performance is clarified, and the optimized experimental parameters are determined.

[0008] The specific steps are as follows:

[0009] (1) Experimental design and parameter analysis

[0010] Through single factor experiments, under the condition of changing a single process parameter, the influence of each process parameter on the coating forming quality is determined; according to the single factor experimental results, the working range of each process parameter is determined; the process parameters include power, bias, working gas pressure and target-substrate distance; the coating forming quality includes density, bonding strength, hardness and microstructure;

[0011] (2) Model fitting and regression model testing

[0012] Based on the working range of each process parameter determined in step (1), a response surface experiment is performed to determine the influence of multiple process parameters on the coating forming quality within the working range of each process parameter; the experimental data obtained by the response surface is used as a database to establish a mathematical model of multiple process parameters and coating forming quality based on data-driven, as follows:

[0013] When there is a linear relationship between the coating forming quality and the process parameters, the mathematical model is expressed using a first-order function model:

[0014] y=θ 0 +θ 1 x 1 +θ 2 x 2 +…+θ k x k+ε (1)

[0015] When there is a nonlinear trend in the mathematical model, that is, there is a nonlinear relationship between the coating forming quality and the process parameters, a high-order polynomial is used to express it, and the mathematical model is expressed using a second-order regression model:

[0016]

[0017] Where y is the predicted coating quality, θ 0 is a constant term, θ i is the linear coefficient, x i is the process parameter variable, θ ii For x i The quadratic coefficient, θ ij For x i and x j The coupling interaction coefficient between them, ε is the residual error term, and k is the number of input variables;

[0018] Then the P value was used to test the fit of the mathematical model for evaluation, with the coating forming quality as the output index and the process parameters as the input variables; when the P value was less than 0.05, it indicated that the mathematical model was effective and the fitting effect was excellent; when the P value was greater than 0.05, it may be due to the influence of data noise, indicating that the mathematical model had no significant effect on the output index, and the mathematical model needed to be optimized, that is, the significance of the linear terms, interaction terms, and quadratic terms in the mathematical model needed to be analyzed; by comparing the P values ​​of the mathematical model, with the goal of establishing the optimal mathematical model with a high degree of fit, the mathematical model was fitted and optimized, taking into account all the input variables, the interaction terms between the input variables, and the square terms; by the stepwise regression method, the insignificant terms with P>0.05 in the mathematical model were eliminated one by one to ensure that each term retained in the final mathematical model had a significant impact on the output index, thereby ensuring that the mathematical regression model had high accuracy and reliability; the correlation between the response surface experimental value of the coating forming quality and the predicted value obtained by the mathematical model was calculated using the corrected multiple error coefficient AdjR 2 To express, AdjR 2 The value range is 0-1. The larger the value, the better the fit between the coating molding quality and the mathematical model obtained by the process parameters, thus achieving the purpose of verifying the accuracy of the mathematical model.

[0019] (3) Determine multi-parameter interactions

[0020] The relationship between process parameters and coating forming quality is intuitively displayed in the form of 3D response surface diagram and contour map, and the influence of process parameters on coating forming quality is analyzed and evaluated; the 3D response surface diagram intuitively shows the influence of any two process parameters on coating forming quality, and the maximum value of coating forming quality is the highest point of the surface; in the contour map, if the shape is circular, it indicates that the interaction effect of the two process parameters on the coating forming quality is weak, and if the shape is elliptical, it indicates that the two process parameters have a significant interaction on the coating forming quality;

[0021] (4) Optimization of coating process parameters based on coating forming quality target value

[0022] Since it is necessary to optimize multiple coating forming qualities and process parameters at the same time, the satisfaction function is used to achieve the coordinated optimization of multiple process parameters. For different optimization directions, the "big", "small" and "eye" functions are used to express them respectively, as shown in formulas (3)-(5). The composite desirability formula (6) is used to achieve the balanced optimization of multiple eigenvalues ​​and thus determine the optimal combination of process parameters by comprehensively considering the optimization requirements of each target.

[0023] Among them, the expected large function: the larger the target value, the better; the expected small function: the smaller the target value, the better; the expected target function: the target value is expected to be close to the target value;

[0024]

[0025] in, is the function prediction value of the i-th coating forming quality, d i is the desirability of the i-th coating quality, U i is the maximum acceptable value of the i-th coating forming quality, D is the composite desirability, T i is the target value of the i-th coating forming quality, r i is the weight of the desirability of the ith output, L i is the minimum acceptable value of the i-th coating feature, w i is the weight of the ith output, and w is ∑w i , n is the number of responses.

[0026] Beneficial effects of the present invention:

[0027] (1) The present invention clarifies the influence of the process parameter target-substrate distance on the geometric and performance characteristics of the coating during magnetron sputtering deposition.

[0028] (2) The present invention establishes a model for predicting the geometric characteristics and performance of magnetron sputtering deposition coatings based on the interaction of multiple process parameters (sputtering power, substrate bias, sputtering gas pressure and target-substrate distance), and conducts experimental verification.

[0029] (3) The present invention has strong applicability and is not limited by the properties, size, etc. of the coating material or the substrate material, and has good adaptability.

[0030] (4) The present invention solves the problems of high process cost, low efficiency and uncontrollable prediction accuracy of the original magnetron sputtering coating geometric feature prediction method based on multiple experiments, laying a foundation for the subsequent establishment of process planning and intelligent manufacturing that can achieve rapid automation in the deposition manufacturing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Schematic diagram of magnetron sputtering principle.

[0032] Figure 2 Schematic diagram of the coating to be cut.

[0033] Figure 3 Schematic diagram of coating geometric features.

[0034] Figure 4 Schematic diagram of the bearing coating structure.

[0035] Figure 5 Flowchart for the prediction of geometric features of magnetron sputtered coatings. DETAILED DESCRIPTION

[0036] The specific implementation of the present invention is further described below in conjunction with the accompanying drawings and technical solutions.

[0037] In order to improve the wear resistance, corrosion resistance and friction reduction performance of the bearing, magnetron sputtering aluminum-tin alloy coating technology can be used, such as Figure 4 As shown in the figure, the aluminum-tin alloy coating can effectively reduce the friction coefficient, reduce heat accumulation, and improve lubrication performance under high load and high-speed operation conditions, thereby extending the service life of the bearing and ensuring its stable operation.

[0038] Theoretically, magnetron sputtering technology can form coatings of different materials, but there are still many problems to be solved in actual production, which limits the application scope of magnetron sputtering technology. At present, there are three main problems in magnetron sputtering technology. The first is the equipment: magnetron sputtering equipment is usually more complex, with high equipment cost and difficulty in maintenance; the second is the material. For aluminum-tin coatings, tin has a low melting point and is easy to volatilize, which leads to tin loss or non-uniform distribution at high temperatures. Controlling the tin content and distribution in deposition is a major problem, which usually needs to be adjusted by optimizing the deposition parameters; the third is the process: process parameters (sputtering power, substrate bias, sputtering gas pressure, target-substrate distance and other process parameters) directly affect the coating morphological quality (geometric dimensions, microhardness, coating adhesion, etc.). With the current equipment, the problem of quickly establishing an accurate coating morphological quality prediction model is affecting the development and application of magnetron sputtering technology and has become a key issue in improving the quality of coating formation.

[0039] In the field of magnetron sputtering aluminum-tin alloy coating, there are currently studies on parameters such as substrate temperature, deposition gas pressure, substrate bias and sputtering power. Figure 1 As shown in the figure, if the target-substrate distance is reduced, the distance that the atoms sputtered from the target move to the substrate becomes shorter, which reduces the probability of collision with the working gas, resulting in more atoms moving to the substrate and a larger coating thickness. However, there is currently no research on the effect of the target-substrate distance parameter on the key morphological quality of the coating. At the same time, the current relationship between the process parameters and the morphological quality of magnetron sputtering aluminum-tin alloy coatings is mostly based on single-factor analysis, lacking the study of the interaction relationship between multiple factors and the exploration of the mapping relationship of morphological quality under complex multi-process coupling, and it is impossible to optimize the process for complex deposition processes. The interaction between multiple processes will bring great changes to the key morphological quality of the coating. The process cost based on multi-experimental data exploration is expensive and inefficient, and the prediction accuracy is uncontrollable. At present, there is no effective method to quickly and accurately establish a prediction model for the geometry and performance characteristics of the coating. Therefore, it is necessary to propose a prediction method for the geometry and performance characteristics of magnetron sputtering aluminum-tin alloy coatings under the action of multi-process coupling, clarify the law of the change of the morphological quality of the coating under the action of process coupling, and optimize the process parameters to complete the rapid and accurate prediction of the geometric characteristics of the coating, so as to obtain a high-performance aluminum-tin alloy coating.

[0040] The present invention proposes a method for predicting the geometry and performance characteristics of magnetron sputtering deposited coatings. The specific process is as follows: Figure 5 As shown, the specific steps of this method are as follows:

[0041] (1) Experimental design and parameter analysis

[0042] Through single factor experiments, under the condition of changing a single process parameter, the influence of each process parameter on the coating forming quality is determined; according to the single factor experimental results, the working range of each process parameter is determined; the process parameters include power, bias, working gas pressure and target-substrate distance; the coating forming quality includes density, bonding strength, hardness and microstructure;

[0043] (2) Model fitting and regression model testing

[0044] Based on the working range of each process parameter determined in step (1), a response surface experiment is performed to determine the influence of multiple process parameters on the coating forming quality within the working range of each process parameter; the experimental data obtained by the response surface is used as a database to establish a mathematical model of multiple process parameters and coating forming quality based on data-driven, as follows:

[0045] When there is a linear relationship between the coating forming quality and the process parameters, the mathematical model is expressed using a first-order function model:

[0046] y=θ 0 +θ 1 x 1 +θ 2 x 2 +…+θ k x k +ε (1)

[0047] When there is a nonlinear trend in the mathematical model, that is, there is a nonlinear relationship between the coating forming quality and the process parameters, a high-order polynomial is used to express it, and the mathematical model is expressed using a second-order regression model:

[0048]

[0049] Where y is the predicted coating quality, θ 0 is a constant term, θ i is the linear coefficient, x i is the process parameter variable, θ ii For x i The quadratic coefficient, θ ij For x i and x j The coupling interaction coefficient between them, ε is the residual error term, and k is the number of input variables;

[0050] Then the P value was used to test the fit of the mathematical model for evaluation, with the coating forming quality as the output index and the process parameters as the input variables; when the P value was less than 0.05, it indicated that the mathematical model was effective and the fitting effect was excellent; when the P value was greater than 0.05, it may be due to the influence of data noise, indicating that the mathematical model had no significant effect on the output index, and the mathematical model needed to be optimized, that is, the significance of the linear terms, interaction terms, and quadratic terms in the mathematical model needed to be analyzed; by comparing the P values ​​of the mathematical model, with the goal of establishing the optimal mathematical model with a high degree of fit, the mathematical model was fitted and optimized, taking into account all the input variables, the interaction terms between the input variables, and the square terms; by the stepwise regression method, the insignificant terms with P>0.05 in the mathematical model were eliminated one by one to ensure that each term retained in the final mathematical model had a significant impact on the output index, thereby ensuring that the mathematical regression model had high accuracy and reliability; the correlation between the response surface experimental value of the coating forming quality and the predicted value obtained by the mathematical model was calculated using the corrected multiple error coefficient AdjR 2 To express, AdjR 2 The value range is 0-1. The larger the value, the better the fit between the coating molding quality and the mathematical model obtained by the process parameters, thus achieving the purpose of verifying the accuracy of the mathematical model.

[0051] (3) Determine multi-parameter interactions

[0052] The relationship between process parameters and coating forming quality is intuitively displayed in the form of 3D response surface diagram and contour map, and the influence of process parameters on coating forming quality is analyzed and evaluated; the 3D response surface diagram intuitively shows the influence of any two process parameters on coating forming quality, and the maximum value of coating forming quality is the highest point of the surface; in the contour map, if the shape is circular, it indicates that the interaction effect of the two process parameters on the coating forming quality is weak, and if the shape is elliptical, it indicates that the two process parameters have a significant interaction on the coating forming quality;

[0053] (4) Optimization of coating process parameters based on coating forming quality target value

[0054] Since it is necessary to optimize multiple coating forming qualities and process parameters at the same time, the satisfaction function is used to achieve the coordinated optimization of multiple process parameters. For different optimization directions, the "big", "small" and "eye" functions are used to express them respectively, as shown in formulas (3)-(5). The composite desirability formula (6) is used to achieve the balanced optimization of multiple eigenvalues ​​and thus determine the optimal combination of process parameters by comprehensively considering the optimization requirements of each target.

[0055] Among them, the expected large function: the larger the target value, the better; the expected small function: the smaller the target value, the better; the expected target function: the target value is expected to be close to the target value;

[0056]

[0057] in, is the function prediction value of the i-th coating forming quality, d i is the desirability of the i-th coating quality, U i is the maximum acceptable value of the i-th coating forming quality, D is the composite desirability, T i is the target value of the i-th coating forming quality, r i is the weight of the desirability of the ith output, L i is the minimum acceptable value of the i-th coating feature, w i is the weight of the ith output, and w is ∑w i , n is the number of responses.

[0058] (5) Experimental verification

[0059] Based on the optimized mathematical model, aluminum-tin alloy coating was prepared by magnetron sputtering, and the geometric and performance characteristics were compared with the predicted results of the mathematical model to verify the effectiveness of the model.

[0060] It is easy for those skilled in the art to understand that any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting the geometry and performance characteristics of magnetron sputtering aluminum-tin coatings, characterized in that: Here are the steps: (1) Experimental design and parameter analysis Through single factor experiments, under the condition of changing a single process parameter, the influence of each process parameter on the coating forming quality is determined; according to the single factor experimental results, the working range of each process parameter is determined; the process parameters include power, bias, working gas pressure and target-substrate distance; the coating forming quality includes density, bonding strength, hardness and microstructure; (2) Model fitting and regression model testing Based on the working range of each process parameter determined in step (1), a response surface experiment is performed to determine the influence of multiple process parameters on the coating forming quality within the working range of each process parameter; the experimental data obtained by the response surface is used as a database to establish a mathematical model of multiple process parameters and coating forming quality based on data-driven, as follows: When there is a linear relationship between the coating forming quality and the process parameters, the mathematical model is expressed using a first-order function model: y=θ0+θ1x1+θ2x2+…+θ k x k +e (1) When there is a nonlinear trend in the mathematical model, that is, there is a nonlinear relationship between the coating forming quality and the process parameters, a high-order polynomial is used to express it, and the mathematical model is expressed using a second-order regression model: Where y is the predicted coating forming quality, θ0 is a constant term, and θ i is the linear coefficient, x i is the process parameter variable, θ ii For x i The quadratic coefficient, θ ij For x i and x j The coupling interaction coefficient between them, ε is the residual error term, and k is the number of input variables, i.e., the number of process parameters; Then the P value is used to test the fit of the mathematical model for evaluation, with the coating forming quality as the output index and the process parameters as the input variables; when the P value is less than 0.05, it indicates that the mathematical model is effective and the fitting effect is excellent; when the P value is greater than 0.05, it may be due to the influence of data noise, indicating that the mathematical model has no significant effect on the output index, and the mathematical model needs to be optimized, that is, the significance of the linear terms, interaction terms, and quadratic terms in the above mathematical model needs to be analyzed; by comparing the P values ​​of the mathematical model, with the goal of establishing the optimal mathematical model with a high degree of fit, the mathematical model is fitted and optimized, taking into account all the input variables, the interaction terms and square terms between the input variables; by the stepwise regression method, the insignificant terms with P>0.05 in the mathematical model are eliminated one by one to ensure that each term retained in the final mathematical model has a significant impact on the output index, thereby ensuring that the mathematical regression model has high accuracy and reliability; the correlation between the experimental value of the coating forming quality response surface and the predicted value obtained by the mathematical model is corrected by the multiple error coefficient AdjR 2 To express, AdjR 2 The value range is 0-1. The larger the value, the better the fit between the coating molding quality and the mathematical model obtained by the process parameters, thus achieving the purpose of verifying the accuracy of the mathematical model. (3) Determine multi-parameter interactions The relationship between process parameters and coating forming quality is intuitively displayed in the form of 3D response surface diagram and contour map, and the influence of process parameters on coating forming quality is analyzed and evaluated; the 3D response surface diagram intuitively shows the influence of any two process parameters on coating forming quality, and the maximum value of coating forming quality is the highest point of the surface; in the contour map, if the shape is circular, it indicates that the interaction effect of the two process parameters on the coating forming quality is weak, and if the shape is elliptical, it indicates that the two process parameters have a significant interaction on the coating forming quality; (4) Optimization of coating process parameters based on coating forming quality target value Since it is necessary to optimize multiple coating forming qualities and process parameters at the same time, the satisfaction function is used to achieve the coordinated optimization of multiple process parameters. For different optimization directions, that is, different target values ​​in coating forming quality, the "big", "small" and "eye" functions are used to express them respectively, as shown in formulas (3)-(5). The composite desirability formula (6) is used to achieve the balanced optimization of multiple eigenvalues ​​and thus determine the optimal combination of process parameters by comprehensively considering the optimization requirements of each target. Among them, the expected large function: the larger the target value, the better; the expected small function: the smaller the target value, the better; Wangmu function: the expected target value is close to the target value; in, is the function prediction value of the i-th coating forming quality, d i is the desirability of the i-th coating quality, U i is the maximum acceptable value of the i-th coating forming quality, D is the composite desirability, T i is the target value of the i-th coating forming quality, r i is the weight of the desirability of the ith output, L i is the minimum acceptable value of the i-th coating feature, w i is the weight of the ith output, and w is ∑w i , n is the number of responses.

Images