Narrow Results

The aim of this research is study the effect of polishing factors to the reduction ratio rate in surface roughness (%$\mathrm{\Delta }Ṙ)$ during the Magnetic Abrasive Finishing (MAF) process using Response Surface Methodology (RSM). The parameters studied were machining gap, rotational speed, abrasive size and magnetic abrasive particle (MAP) size. Quadratic models were developed by applying Box–Behnken Design (BBD). Also, experiments were carried out on the silicon wafer and results of surface roughness data were analyzed by using analysis of variance (ANOVA) and significant factors were identified. According to our findings, the maximum %$\mathrm{\Delta }Ṙ$ value and the best surface roughness of silicon wafer achieved 3.70 and 51 nm, respectively.

The effect of various laser processing parameters on the kerf width and cut quality of Si wafer as well as encapsulated Si wafer is investigated. The parameters are then optimized to minimize the heat affect zone and obtain the best possible cut quality. It has been found that oxygen is the most suitable assist gas for laser dicing and that the highest gas pressure may not produce the best cut quality. The effect of laser repetition rate, pump energy, feed rate, and number of passes are also studied. Under optimized parameters, the cut quality of Si wafer using laser dicing is found to be comparable to diamond saw dicing.

This work illustrates how the separation of a semiconductor wafer into individual devices occurs during conventional mechanical dicing. In situ examinations indicate that the final separation of the wafer takes place before the dicing blade has fully penetrated its active surface. Thus, it was predicted that mechanical dicing-induced damage in the separated device patterns would be due to other mechanical actions rather than the grinding action between the diamond particles embedded in the blade and the wafer. Based on the in situ examinations, it was experimentally tested how manipulating the revolving speed of the dicing blade affected the prevention of dicing-induced damage to device patterns. The experimental results show that among various mechanical actions, the impact stress due to the revolving action of the blade could be the most possible candidate for damage in the device pattern on the final uncut semiconductor wafer.

This work presents the feasibility of picosecond laser micromachining of polysilicon wafer. Surface topography, microstructure and residual stress of both as-received surface and laser-machined surface were analyzed carefully by confocal microscope, scanning electron microscope and Raman microscope. Moreover, electrical properties of laser-machined wafer have been investigated to examine the effect of laser micromachining on Si substrate via characterizations of resistivity and $I-V$ curves. The results show that the wafer thickness has been reduced up to 50%, while the depth of HAZ is less than 3$\mu$m, and compressive stress can be achieved at the laser-machined surface. Besides, laser micromachining causes little influence on electrical properties of wafer. This proof-of-concept process has the potential application in mass production of integrated circuit industry.