Single-phase string inverter has been widely applied to grid-tied applications due to its efficient integration of renewable energies. However, conventional approaches of inverter loss estimation oversimplify switching loss calculations by assuming hard switching exclusively, thereby neglecting the effects of soft-switching and partial soft switching modes during sinusoidal load variations, as well as conduction losses during dead time. To address these limitations, this paper presents a comprehensive analysis of switching modes in SiC-based inverters. Furthermore, the switching process stages are redefined through energy transfer dynamics between the channel and junction capacitances. Based on this, an analytical model for SiC MOSFETs switching loss is proposed through time-domain differential equations. Additionally, this paper introduces a loss model-based dead time algorithm aimed to figure out the optimal dead time with reduced power loss and improved THD performance. The analysis also details how dead time influences switching modes. Finally, A 1-kW prototype is established to proof.