CT-FFR based on alternative boundary conditions and reduced-order fluid model is feasible, highly reproducible, and may be accurate in detecting FFR ≤ 0.8. It requires a short processing time and can be completed at point-of-care. Further validation is required in large prospective multicenter settings.

Bipolar degradation, which is caused by the expansion of stacking faults (SFs) during operation, has been a serious issue in 4H-SiC power devices. To evaluate the threshold minority carrier density of SF expansion, ρth, Maeda et al. proposed a theoretical model based on quantum well action and dislocation theory. This model includes SF energy variations, electronic energy lowering due to carrier trapping, and resolved shear stress applied to partial dislocations, τrss. Though the SF energy and the electric energy lowering were quantitatively established, the effect of τrss has not been discussed well yet. In this study, we first conducted theoretical predictions of the effect of τrssonρth. Then, based on our previous experiment on the dependence of threshold current density on mechanical external stress, we investigated the dependence of ρthonτrss. We conducted submodeling finite element analysis to obtain τrss induced by both residual stress due to the fabrication process and experimentally applied external stress. Finally, we obtained ρth at the origin of SF expansion from the experimentally measured threshold current density using device simulation. It was found that the dependence of ρthonτrss was almost linear. Its gradient was −0.04 ± 0.01 × 1016 cm−3/MPa, which well agrees with the theoretical prediction of −0.03 ± 0.02 × 1016 cm−3/MPa. Our study makes possible a comprehensive evaluation of the critical condition of bipolar degradation.

Expansion of a single Shockley stacking fault (SSF) during forward-current operation decreases the reliability of 4H-SiC bipolar devices. We propose a practical method for analyzing the defect evolution of SSF expansion based on free energy according to current density, temperature, and resolved shear stress conditions. The free energy includes chemical potential and elastic strain energy. Specifically, the chemical potential is related to the driving force for the formation of SSFs by temperature and current, and the elastic strain energy corresponds to the driving force for dislocations that form SSFs under the applied stress. It was confirmed that the proposed multiphysics method could well simulate SSF evolution when stress and current were applied. Furthermore, the results suggest that quantum well action, in which electrons in n-type 4H-SiC enter SSF-induced quantum well states to lower the energy of the dislocation system, affects the driving force of SSF formation.

No abstract

No abstract

Single Shockley stacking faults (SSFs) expand from basal plane dislocations (BPDs) under forward current operation of 4H-SiC bipolar devices, giving rise to a reliability deterioration mode called “bipolar degradation”. Several groups have proposed models for the expansion of SSFs, in which the SSFs expand when electron-hole pair recombination takes place at BPDs. Maeda proposed a formulation of SSF expansion that includes stacking fault energy. However, the mechanisms by which mechanical stress affects the expansion of SSFs are unclear. In this paper, we evaluated the “expansion threshold current” of bar-shaped SSFs in a mechanical stress field using a p-i-n diode fabricated on 4H-SiC. To confirm the effect of mechanical stress on the threshold current for bar-shaped SSF expansion, a SiC-p-i-n diode was evaluated by the four-point bending method. Experimental results show that the threshold current of SSFs decreases or increases by more than 100 A/cm2 depending on the direction of the applied stress of SSFs. This result indicates that mechanical stress is an important factor for SiC bipolar device design.

We detected all components of the deformation potential constants of 4H-SiC by first-principles calculations and developed a method to estimate the stress distribution in 4H-SiC power devices by Finite Element Method (FEM) and Raman spectroscopy. The values of bA1, aE2, and bE2 obtained by calculations agreed well with experimental results, while those of aA1, bE1, and cE2 were about 45% larger. The relationship between phonon frequency and stress was nonlinear as shear stress increased. The multistep FEM analysis reproducing the manufacturing process is also conducted. The stress distribution was converted to the Raman shift and compared with results of micro-Raman spectroscopy. Except for the interface between SiC and the electrode, the analysis results agreed well with the experimental results. It was found that a compressive stress of about 200 MPa at the SiC/electrode interface and a resolved shear stress of about 20 MPa at the epilayer/substrate interface were generated.

No abstract