Influence of Field Plate, Gate Width, and Voltage Dependence of Thermal Resistance (<i>R</i>TH) for AlGaN/GaN HEMT

The thermal resistance (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {TH}}$ </tex-math></inline-formula>) of 0.15- and 0.25-<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $ </tex-math></inline-formula> m AlGaN/GaN high-electron mobility transistor (HEMT) is computed from the linear portion of simulated change in channel temperature (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\Delta {T}$ </tex-math></inline-formula>) versus power dissipation (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${P}_{D}$ </tex-math></inline-formula>) plot. The simulated I–V characteristics are validated with measured data using the Sentaurus thermodynamic TCAD model. Subsequently, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {TH}}$ </tex-math></inline-formula> is estimated for the HEMT with and without a gate field-plate (FP) structure. Three linear <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {TH}}$ </tex-math></inline-formula> regions are observed for HEMT with FP, due to the variation in the hotspot position with the increasing drain voltage, whereas a single <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {TH}}$ </tex-math></inline-formula> is identified for the HEMT without FP, as the hotspot remains at the drain side of the gate edge over a wide range of dissipated power. <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {TH}}$ </tex-math></inline-formula> is also reported for the HEMT with different FP lengths (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${L}_{\text {FP}}$ </tex-math></inline-formula>), showing that <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {TH}}$ </tex-math></inline-formula> decreases as <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${L}_{\text {FP}}$ </tex-math></inline-formula> increases. <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {TH}}$ </tex-math></inline-formula> is found to decrease with the gate width, which is consistent with the theory as well as the thermoreflectance and pulsed I– V measurements. A slight increase in <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {TH}}$ </tex-math></inline-formula> is observed upon reducing the gate voltage. Furthermore, the change in <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {TH}}$ </tex-math></inline-formula> is evaluated by placing thermal contact at three different locations (substrate bottom, buffer and substrate sidewall, and contacts) in the simulation. Similar observations are detected for 0.25-<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $ </tex-math></inline-formula> m AlGaN/GaN HEMTs.