Advanced Semiconductor Laboratory
Developing cutting-edge technologies based on III-nitride semiconductors

Blog Posts

Doping of AlN and its alloy

Kuang-Hui Li

April 26, 2017 - Posted in Discussion
Doping of semiconductor is very crucial since pure semiconductor material couldn’t make into a device. Doping of nitride material is not trivial like other III-V material, and it is not fully understood yet. AlN is the hardest compared to GaN and InN.
The reason why the dopants don’t incorporate into AlN and high-Al-content AlGaN might be the high activation energy requirement.
As shown in Fig. 3, the activation energy of Si (n-dopant) increases drastically when Al content increases in AlGaN, and reach to the highest point ~280 meV in pure AlN. If the activation energy of Si is high, only a few percent of Si is activated in AlN and AlGaN. Could we increase the concentration of Si? The answer is NO. There is a self-compensate effect if the concentration of Si is too high, as shown in Fig. 4. High Si concentration may compromise the crystal quality which increases the dislocation density and Al- or N-vacancy. It is even challenging for p-doping. The activation energy of Mg (p-dopant) is 510 meV in AlN.
AlN Si 1.png
AlN Si 2.png
It has been known that E-beam and thermal annealing can activate dopant in GaN, InN, and their alloy. However, these methods don’t work well in AlN. If E-beam can activate dopant in GaN, could we develop an in-situ activation source during the growth of AlN? Maybe we can expose the substrate to UV, X-ray, or even gamma-ray during the growth. Put the safety issues aside, how to design a system with in-situ activation source?

Reference: Doping of III-nitride materials, Materials Science in Semiconductor Processing 62 (2017) 180–191
Leave a Comment
* Comment: