Analytical prediction for depth of subsurface damage in silicon wafer due to self-rotating grinding process

Subsurface damage (SSD) induced by silicon wafer grinding process is an unavoidable problem in semiconductor manufacturing. Although experimental attempts have been made on investigation of the influential factors on the SSD depth, however, few theoretical studies have been conducted to obtain SSD depth through grinding parameters. To fill the gap, an analytical model is developed to predict the SSD depth in silicon wafer due to self-rotating grinding process, which can reveal the relationship among SSD depth and the grinding parameters, the size of the abrasive grains and the radial distance from the wafer center. The establishment of the proposed model is based on scratch theory and fracture mechanics of isotropic brittle materials, and we further consider the effects of elastic recovery, cleavage plane and crystalline orientation on SSD formation. To validate the applicability of the proposed predictive model, grinding experiments with varied grinding parameters are performed and the depths of SSD along the <110> and <100> crystal directions are also measured and analyzed. The results given by the proposed model present reasonable accuracy of less than 20% deviation with experimental results. Effects of grinding parameters, wafer radial distance, crystalline orientation, and abrasive grain size on SSD depth are discussed in detail.