Workpiece rotational grinding is widely used in the ultra-precision machining of hard and brittle semiconductor materials, including single-crystal silicon, silicon carbide, and gallium arsenide. Surface roughness and subsurface damage depth (SDD) are crucial indicators for evaluating the surface quality of these materials after grinding. Existing prediction models lack general applicability and do not accurately account for the complex material behavior under grinding conditions. This paper introduces novel models for predicting both surface roughness and SDD in hard and brittle semiconductor materials. The surface roughness model uniquely incorporates the material’s elastic recovery properties, revealing the significant impact of these properties on prediction accuracy. The SDD model is distinguished by its analysis of the interactions between abrasive grits and the workpiece, as well as the mechanisms governing stress-induced damage evolution. The surface roughness model and SDD model both establish a stable relationship with the grit depth of cut (GDC). Additionally, we have developed an analytical relationship between the GDC and grinding process parameters. This, in turn, enables the establishment of an analytical framework for predicting surface roughness and SDD based on grinding process parameters, which cannot be achieved by previous models. The models were validated through systematic experiments on three different semiconductor materials, demonstrating excellent agreement with experimental data, with prediction errors of 6.3% for surface roughness and 6.9% for SDD. Additionally, this study identifies variations in elastic recovery and material plasticity as critical factors influencing surface roughness and SDD across different materials. These findings significantly advance the accuracy of predictive models and broaden their applicability for grinding hard and brittle semiconductor materials.

Silicon carbide (SiC) has a wide range of application prospects for the excellent characteristics. However, its high hardness, brittleness, and chemical inertia improve the processing difficulty, which restricts the popularization and application of single-crystal SiC semiconductor devices. This paper introduces the research progress of SiC from two parts: material removal mechanism and ultraprecision machining technology. The material removal and damage formation mechanism of SiC at home and abroad, as well as the research progress of lapping, polishing technology and ultraprecision grinding technology are introduced in detail. The analysis shows that there are some differences in the removal mechanisms of SiC studied by different scholars. In addition, the lack of a reasonable theoretical model for surface integrity hinders the selection of efficient and low-damage process parameters for grinding SiC wafers. In terms of single crystal SiC ultraprecision machining, the more mature machining methods at this stage mainly go through three steps: double-sided lapping, single-sided lapping and chemical mechanical polishing. The machining efficiency and surface integrity of each step affect the production efficiency and scrap rate of the final product. As SiC wafers develop towards larger sizes, ultraprecision grinding technology, which utilizes workpiece rotation grinding principles, emerges as an efficient and low-damage machining method of SiC wafers, has the potential to replace the traditional lapping.

The roughness of the contact surface exerts a vital role in rubbing. It is still a significant challenge to understand the microscopic contact of the rough surface at the atomic level. Herein, the rough surface with a special root mean square (RMS) value is constructed by multivariate Weierstrass–Mandelbrot (W–M) function and the rubbing process during that the chemical mechanical polishing (CMP) process of diamond is mimicked utilizing the reactive force field molecular dynamics (ReaxFF MD) simulation. It is found that the contact area A/A₀ is positively related with the load, and the friction force F depends on the number of interfacial bridge bonds. Increasing the surface roughness will increase the friction force and friction coefficient. The model with low roughness and high lubrication has less friction force, and the presence of polishing liquid molecules can decrease the friction force and friction coefficient. The RMS value and the degree of damage show a functional relationship with the applied load and lubrication, i.e., the RMS value decreases more under larger load and higher lubrication, and the diamond substrate occurs severer damage under larger load and lower lubrication. This work will generate fresh insight into the understanding of the microscopic contact of the rough surface at the atomic level.