Discipline: Technology and Engineering
Subcategory: Electrical Engineering
Sebastian O. Manzo - Pennsylvania State University
Molybdenum disulfide makes for a good semiconductor in transistors because it has a high carrier mobility, a high on/off ratio and robust mechanical properties. There are established thinning methods that could be applied to thinning MoS2, but have not shown the desired control and consistency. Consequently, this research project attempted to test and improve these known etching and exfoliating techniques, so that more effective thinning processes for MoS2 can be found. Initially, ultrasonic and adhesive tape exfoliation was used. The tape was used to strip layers of molybdenite mineral and those flakes could be increasingly thinned by using more tape. For ultrasonic exfoliation, water and isopropyl alcohol served as the sonicating liquids. Material was scraped off the molybdenite mineral with a razor and dropped in a beaker containing the sonicating liquid. The mixture would then be exfoliated using a bath sonicator and probe sonicator that were used in conjunction with each other. Wet etching was done by filling a beaker with 68% HNO3 or 33-40% HCl and heating the acid on a hot plate. The silicon substrate with MoS2 flakes would float on the acidic solution for a certain amount of time. The acid temperature and etching time was varied in the experiment. Reactive ion etching (RIE) was also tested, using a combination of Ar and NF3 as etchants. Different ratios of Ar to NF3 were used and the self-bias and power were changed, while the pressure remained constant at 150 mT. The results were mostly ineffective and either were too aggressive or showed no signs of etching. Adhesive tape exfoliation proved to yield more and thinner MoS2 flakes than ultrasonic. The ultrasonic-exfoliated samples were left with too much residue after the mixture evaporated, making it difficult to see any flakes at all on the microscope. Those that were visible appeared to be significantly thicker. Only one sample was successfully wet etched with HCl at 50C and 1 hour etching time. However, it was not uniform and only the smaller flakes seemed to be affected. All other trials with HCl and HNO3 at different temperatures and etching times showed no etching of the flakes at all. RIE was the most successful method, since it was found to etch MoS2 easily. Progress was made in finding a serviceable recipe that etched the flakes in a more controlled fashion. At this point, these etching methods are not reliable. More research has to go into the effect of different acids as etchants. Aqua Regia or H202+HCl are recipes worth trying, since they were used to remove MoS2 lubricant in old literature. RIE seems to be the most promising method, but more experimentation is required so that the etch is more controlled. An atomic force microscope would be essential to have so that the exact flake thickness is known. Knowing the exact thickness would help determine the time required to etch one layer at a time, and would give more insight on which parameters to adjust.
Funder Acknowledgement(s): This study could not have been accomplished without the support of the National Science Foundation grant for the Emerging Frontiers in Research and Innovation (EFRI) for Two-Dimensional Atomic Layer Research and Engineering (2-DARE). I would also like to acknowledge Dr. Jackson and Yiyang Gong for the help they provided.
Faculty Advisor: Yiyang Gong, ywg5045@psu.edu