A Machine Learning-Assisted Model for GaN Ohmic Contacts Regarding the Fabrication Processes
ABSTARCT :
Gallium nitride (GaN) devices have been successfully commercialized due to their superior performance, especially their high-power transformation efficiency. To further reduce the power consumption of these devices, the optimization for the ohmic contacts is attracting more and more attention. In the light of the mature and powerful machine learning (ML) techniques, this work provides a novel method to evaluate the fabrication processes of the ohmic contacts in AlGaN/GaN heterojunction, n-type, and p-type GaN, by establishing a regression-based model. The proposed model can not only investigate the influence weight of each process but also predict the contact resistance by inputting the desired recipes. A website ( http://ohmic.zeheng.wang/ ) containing the successfully trained model for the readers’ interests is also provided, which, we believe, would benefit the society of the process development and optimization.
EXISTING SYSTEM :
AlGaN/GaN HEMT is normally-on (depletion mode) in nature due to the existing 2DEG channel which is induced by the strong polarization charges from the AlGaN/GaN heterostructure. A normally-off (enhancement mode) device is desirable as it is able to eliminate negative power supply which reduces the overall design cost, size and complexity of the system. Several approaches in achieving Enhancement mode HEMT have been reported; such as gate recess etching, fluorine plasma treatment, and AlGaN barrier thinning. However, a standard processing method with good controllability, uniformity and reproducibility is still lacking.
DISADVANTAGE :
These disadvantages of SiC play in favor of silicon as a substrate of choice for GaN epitaxy with its good thermal conductivity, low cost and availability of larger wafers, in addition to the possibility of integrating GaN devices with mature silicon devices.
? On the other hand, the high thermal expansion coefficient mismatch and large lattice mismatch between Si and GaN favor the appearance of cracks on the surface.
PROPOSED SYSTEM :
The proposed treatment requires an additional high temperature (>700 °C) pre-gate annealing using RTP which has to been done prior to gate metal deposition to avoid the degradation of the Schottky gate. As a result, the gate recess etching and gate patterning steps are separated, leading to subsequent lithography and self-aligning issues.Germanium, which acts as an n-type donor in the AlGaN barrier layer, is proposed to be able to compensate the effects of positive sheet charges at the heterointerface, which eventually depletes the 2DEG channel leading to a positive shift in the threshold voltage. The efficiency and feasibility of this approach in achieving an enhancement mode of operation are studied.
ADVANTAGE :
? To understand the superiority of GaN technology over conventional devices at high temperatures, it is very important to identify the main factors that affect the physical limits and operation of semiconductors at high temperature.
? This advantage in wide bandgap devices reduces the junction leakage current by several orders of magnitude at any operating temperature, thus enabling the implementation of devices operating at high voltage and high temperature where conventional semiconductors are no longer applicable.
|