p-GaN Doping Profile for Mg Diffusion Control in HEMTs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing high electron mobility transistors (HEMTs) face challenges in controlling the out-diffusion and activation of magnesium (Mg) or other dopants in the p-type gallium nitride (GaN) layer, leading to performance degradation due to dopant diffusion into the aluminum gallium nitride (AlGaN) layer, resulting in higher on-resistance and high temperature reverse bias (HTRB) issues.
Innovation Solution
The HEMT structure is modified by dividing the p-GaN layer into a main layer and an interface layer, with a lower doping concentration in the interface layer to reduce Mg out-diffusion, and controlling the Mg doping flow using a parabolic function to maintain effective dopant activation, ensuring sufficient activation without increasing total dopant amount.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a Mg-doped p-GaN layer is used to control HEMT switching, then the HEMT can be turned on and off effectively, but the Mg dopant out-diffuses into the AlGaN layer causing performance degradation
Solution Approach 1:
The p-GaN layer is divided into two distinct layers: a first p-GaN layer with higher Mg doping concentration for effective switching control, and a second p-GaN layer with lower Mg doping concentration positioned between the AlGaN layer and the first p-GaN layer to prevent dopant out-diffusion. This segmentation allows each layer to perform its specific function while resolving the contradiction between switching control and performance stability.
Solution Approach 2:
Different regions of the p-GaN structure are assigned different doping concentrations tailored to their specific functions. The first p-GaN layer under the gate contact has high doping for effective switching, while the second p-GaN layer at the AlGaN interface has low doping to prevent out-diffusion. This local quality differentiation resolves the contradiction by optimizing each region for its specific role.
2Ease of operation
If Mg doping concentration is increased to improve switching control, then HEMT on/off capability improves, but dopant out-diffusion into AlGaN layer increases
Solution Approach 1:
The p-GaN layer is segmented into a first p-GaN layer with high Mg doping concentration for switching control and a second p-GaN layer with low Mg doping concentration to prevent out-diffusion. This segmentation allows high doping where needed for switching while preventing harmful out-diffusion at the AlGaN interface.
Solution Approach 2:
The second p-GaN layer acts as an intermediary buffer layer between the AlGaN layer and the first p-GaN layer. This intermediary layer with lower doping concentration prevents direct contact and out-diffusion between high-Mg regions and the AlGaN layer, thereby eliminating the harmful effect while preserving switching control capability.
3Quantity of substance
If dopant out-diffusion occurs in the AlGaN layer, then trapping centers are formed, but this reduces the 2DEG current and increases on-resistance
Solution Approach 1:
The second p-GaN layer serves as a protective intermediary layer that prevents Mg dopant from reaching and forming trapping centers in the AlGaN layer. By blocking the diffusion path, this intermediary layer preserves the 2DEG current and prevents on-resistance increase while allowing the first p-GaN layer to maintain effective switching control.
4Reliability
If total dopant amount is increased to maintain activation, then dopant out-diffusion increases, but activation must be maintained
Solution Approach 1:
The p-GaN layer is segmented into two layers with different doping concentrations. The first p-GaN layer has high Mg doping for effective switching control and activation, while the second p-GaN layer has low doping to prevent out-diffusion. This segmentation maintains necessary activation in the switching region without increasing total dopant amount that would cause harmful out-diffusion.
Solution Approach 2:
Different doping concentrations are applied locally to different regions: high doping in the first p-GaN layer for activation and switching control, and low doping in the second p-GaN layer to prevent out-diffusion. This local quality differentiation maintains activation where needed while preventing harmful effects.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively reduces Mg out-diffusion into the AlGaN layer, eliminating trapping centers and improving HEMT performance by maintaining low on-resistance and preventing HTRB degradation.
Implementation Method 1
A doping concentration of the dopant in the interface layer is lower than that in the main layer
Data Source
AI summary
Structures and methods for controlling dopant diffusion and activation are disclosed. In one example, a semiconductor structure is disclosed. The semiconductor structure includes: a channel layer; a barrier layer over the channel layer; a gate electrode over the barrier layer; and a doped layer formed between the barrier layer and the gate electrode. The doped layer includes (a) an interface layer in contact with the barrier layer and (b) a main layer between the interface layer and the gate electrode. The doped layer comprises a dopant whose doping concentration in the interface layer is lower than that in the main layer.