Artificial Neural Network for Composite Hardness Modeling of Cu/Si Systems Fabricated Using Various Electrodeposition Parameters

Abstract

Copper coatings are produced on silicon wafer by electrodeposition (ED) for various cathode current densities. The resulting composite systems consist of 10 mu m monolavered copper films electrodeposited from sulphate bath on Si wafers with sputtered layers of Cr/Au. Hardness measurements were performed to evaluate properties of the composites. The composite hardness (H-c) was characterized using Vickers microindentation test. Then, an artificial neural network (ANN) model was used to study the relationship between the parameters of metallic composite and their hardness. Two experimental values: applied load during indentation test and current density during the ED process were used as the inputs to the neural network. Finally, the results of the composite hardness (experimental and predicted) were used to estimate the film hardness (H-f) of copper for each variations of the current density. This article shows that ANN is an useful tool in modeling composite hardness change with variation of experimental parameters predicting hardness change of composite Si/Cu with average error of 6 %. Using created ANN model it is possible to predict microhardness of Cu film for current density or indentation load for which we do not have experimental data.