III-N devices have become critical in the fields of energy-efficient lighting and power electronics. Though III-N devices have different materials requirements depending on the end application of the device, one thing common to all devices is the need for extremely pure films. Impurities originating from the atmosphere, wafer processing, or source materials are detrimental both to optoelectronic and transistor devices. This work investigates several impurities found in III-N films grown by plasma assisted molecular beam epitaxy (PAMBE) and analyzes attempts to prevent their incorporation into films.

First Ca was investigated as a potential impurity in PAMBE films following its discovery as a potent impurity in NH3-Assisted MBE materials, especially the critical InGaN active region in light-emitting diodes (LEDs). Although measured Ca levels in PAMBE films was found to be much lower than in NH3-Assisted MBE films, it is still potentially high enough to create a nonradiative recombination center in optoelectronics via Shockley-Read-Hall (SRH) or trap-assisted Auger (TAAR) mechanisms. Growth temperature, the presence or absence of a Ga-adlayer, and Ga-polishing had little effect on Ca incorporation in Ga-polar films. Growth temperature did have a moderate effect on Ca-incorporation into N-polar films, providing a potential path forward for sequestering Ca in lower layers of future devices away from critical device areas.

Next, impurity incorporation in devices requiring a regrowth was examined with an emphasis on Si which is difficult to remove using cleaning procedures employed in MBE growth. The potential for H-radicals to remove Si from a heated wafer surface prior to a regrowth is examined. Though no Si was removed via the procedure used in this work, there exists the possibility that a higher flux of H-radicals, or possibly the use of O-radicals could still be effective in achieving this goal. InN was used as a sacrificial capping layer on GaN films prior to removal from the growth chamber to protect the critical regrowth interface from being exposed to Si and other impurities. This InN cap was not successful at preventing Si from accumulating on the GaN interface, but could be promising as a sacrificial layer to protect the device from processing-related damage prior to regrowth.

Finally, the growth of InGaN films via PAMBE was explored through the use of a modern, high-flux N plasma unit. This plasma unit has been shown to grow smooth GaN films at growth rates over 7 µm/hr. Using this plasma unit capable of over an order of magnitude higher nitrogen flux than previously available led to the growth of InGaN films at a maximum growth rate of 1.3 µm/hr, which is faster than previously recorded in the literature. This enabled more In incorporation into films at higher temperatures due to the stabilization In-N bonds. An InxGa1-xN film with x = 0.05 was grown at 700 °C, which is the highest growth temperature reported for InGaN in the literature. Increasing the growth temperature range and growth rate of InGaN films is likely to lead to smoother films as well as less impurity incorporation which could finally lead to efficient InGaN-based LEDs grown by PAMBE.