Statistical Modeling and Surface Texture Study of Polished Silicon Wafer Si (100) using Chemically Assisted Double Disk Magnetic Abrasive Finishing

The present paper aims to analyze the surface finish of polished silicon wafer by chemically etching with potassium hydroxide (KOH) and mechanical polishing by Double Disk Magnetic Abrasive Finishing (DDMAF) process. The study is emphasized to study the effect of process parameters i.e. polishing speed, working gap, abrasive mesh number and percentage weight of KOH on the surface roughness (R-a). Response surface methodology (RSM) has been used to plan the experiments and Analysis of variance (ANOVA) has been used to analyze the impact of each process parameter on surface roughness. Regression equation for surface roughness in terms of significant process parameters has been developed to determine the surface roughness of polished silicon wafer. The equation has been further optimized using optimizer available with Minitab 17 and Genetic Algorithm (GA) tool box available with MatLab 16, to obtain the optimum process parameters and to prefigure the minimum surface roughness. The confirmatory experiment was carried out at optimum parameters and the prefigured results were found to be closely matched with the experimental findings. The communication further vocalizes the study of surface integrity for the unpolished sample to polished sample at optimum parameter using SEM and AFM images.