STUDY ON FRACTURE TOUGHNESS OF SEMICONDUCTOR MATERIAL USING VICKERS AND BERKOVICH INDENTERS

The indentation method is one of the commonly used methods to determine fracture toughness (K_\rm IC) of brittle materials. One of the challenges is to obtain a suitable equation of the materials from various equations according to different materials and indenters. Therefore, fracture toughness tests with pyramid indenters (Vickers indenter and Berkovich indenter) were conducted on Si (111) and 4H-SiC (0001) under various loads. The crack length c generated in the Vickers indentation experiments were statistically analyzed, and thirteen equations were selected to calculate the fracture toughness of semiconductor materials at room temperature. The applicability of the indentation test was evaluated, based on a comparative analysis with the results of the scratch test. The results show that to eliminate the inherent discreteness of crack length c generated in the Vickers indentation experiment, multiple indentation tests (at least thirty tests) need to be conducted. The ratio of crack length c over the indentation diagonal length a increases with an increase in the applied load P. The crack types of the materials depend on P: Palmqvist crack system appears for low loads and Median crack system appears for high loads. Compared with the average fracture toughness (0.96~MPa,\cdot,\sqrt\rm m and 2.89~MPa,\cdot,\sqrt\rm m, respectively) of Si (111) and 4H-SiC (0001) obtained by micro scratch test, based on linear elastic fracture mechanics (LEFM), the appropriate equations was obtained for both Vickers and Berkovich indenters for the same as material, but which can not be obtained for both Si (111) and 4H-SiC (0001) under the same as indenter from thirteen equations. The fracture toughness of semiconductor materials are best calculated by an expression develope from the Median crack system, and the relationship between fracture toughness being obtained with Vickers indenter and that of with Berkovich indenter is not theoretically 1.073 times, which should be 1.13\pm 0.01.