Date of Graduation
Bachelor of Science in Mechanical Engineering
In the world of semiconductors today, there is a large dissonance between optical devices and electrical application. Due to the limitations of electron transport, photonic integrated circuits are soon-to-be vital in fields like telecommunications and sensing. Right now, these PIC’s are mostly made from indium phosphide. Due to its ubiquitous nature, however, there is a huge push to integrate efficient optics with silicon. It’s cheap, abundant, dope-able, and our electronic infrastructure is based on it. The reason why silicon photonics aren’t already commercialized is because of silicon’s indirect bandgap—it is inefficient with optical applications. The problem with combining direct gap materials like gallium arsenide with silicon comes from the lattice sizes. Almost all commercial direct-gap semiconductors have crystal sizes significantly larger than silicon. When they are grown on top of each other, a lot of strain is introduced at the junction. This causes defects to form, significantly degrading optical efficiency. One solution relies on shrinking the crystal lattice size of a direct bandgap material by alloying. In this project specifically, Boron is being added to Gallium Arsenide. The literature for the binary compound BAs, as well as the ternary alloy BGaAs, is relatively unexplored. The theoretical and experimental data currently available is widely scattered. This inconsistency is due to the challenges facing the growth of BGaAs, but recent advances have made possible to characterization of this fundamentally unknown material.
BGaAs samples were grown via molecular beam epitaxy with boron concentrations ranging from 4%-12%. For characterization, various wavelength dependent measurements were used, including photoluminescence, spectroscopic ellipsometry, UV-Vis and FTIR spectroscopy. Absorption spectra can be indicative of band gap energy, as the material absorbs light at its bandgap energy and higher. From the absorption spectra the data can be manipulated into a Tauc plot to give a quantitative estimation of the bandgap energy, as well as indicate a direct or indirect bandgap. The results are promising. The absorption spectrum shows absorption at wavelengths shorter than 830 nm. The Tauc plot shows a linear region (using direct bandgap parameters), confirming the presence of direct gap recombination. The estimated bandgaps increase as compared to gallium arsenide, which leads to the conclusion of a direct bandgap that increases with increasing boron.
This project lays a foundation for future work. While the maximum concentration tested was only 12% (lattice matching concentration is 24%), the methods and procedure for testing these larger concentrations are laid out. Successful experimental setups and an initial trend is established. More coherent samples should be used for photoluminescence and ellipsometry measurements, as both of these methods require defect free samples. This coherency can be accomplished by thinner samples, and higher levels of boron. Eventually BGaAs can be lattice matched with silicon, and further research can be done with BGaInAs allowing for bandgap control and tunability. Continuing to explore BGaAs is vital, as it shows great promise as a bridge to silicon optical circuits.
McArthur, Joshua, "BGaAs Alloy Semiconductors for Lasers on Silicon" (2019). Mechanical Engineering Undergraduate Honors Theses. 78.
Available for download on Saturday, April 23, 2022