Process-based IGBT Modeling with TCAD for Static and Dynamic Parameters

With the global development in renewable energy, the significance of Insulated Gate Bipolar Transistors (IGBTs) has become increasingly prominent. However, the process-based modeling of IGBT faces problems of low accuracy and low computational efficiency. In this study, a high-precision IGBT model based on the actual processes is established based on the TCAD. Firstly, the key electrical parameters of the selected IGBT device were extracted, including $VGE(th), VCES, VCE( sat ), Cᵢₛₛ, Cₒₛₛ, Cᵣₛₛ, td(off)$, and $tf$. Subsequently, a TCAD model was developed for the device's active area, and calibrated with the Scanning Electron Microscope (SEM) images of actual devices. Furthermore, donor defects and fixed charges were set at the trench interface to match the simulated electrical parameters with measured ones, employing physical models of OldSltboom, Shockley-Read-Hall (SRH), Lackner, Lattice/Carrier Temperature model, etc. Finally, the electrical parameters of the model were calibrated with the experimental results. The error rates of the key static electrical parameters (VGE(th), VCES, VCE (sat)) were controlled within 5%, while the error rates of the capacitor parameters (Cᵢₛₛ, Cₒₛₛ, Cᵣₛₛ) and the switching characteristics (td(off), tf) were controlled within 6% and 12% respectively. Such accuracy of process-based IGBT modeling helps to establish the Safe Operating Area (SOA) of IGBTs and failure mechanism analysis, thus economizing on the development costs.